proceedings of a benthic habitat classification workshop ... · in this report may be factually...

94
C S A S Canadian Science Advisory Secretariat S C C S Secrétariat canadien de consultation scientifique © Her Majesty the Queen in Right of Canada, 2002 © Sa majesté la Reine, Chef du Canada, 2002 ISSN 1701-1280 www.dfo-mpo.gc.ca/csas/ Proceedings Series 2002/023 Série des compte rendus 2002/023 Proceedings of a Benthic Habitat Classification Workshop Compte rendu d’une Atelier sur l’habitat benthique Meeting of the Maritimes Regional Advisory Process Réunion du Processus consultatif des provinces Maritimes Maintenance of the Diversity of Ecosystem Types Maintien de la diversité des types d'écosystème A Framework for the Conservation of Benthic Communities of the Scotian- Fundy Area of the Maritimes Region Cadre de conservation des communautés benthiques du secteur de Scotia-Fundy, Région des Maritimes 25 and 26 June 2001 Les 25 et 26 juin 2001 Bedford Institute of Oceanography Dartmouth, N.S. L’Institut océanographique de Bedford Dartmouth, (N.-É.). J. Arbour (Chair) J. Arbour (président) Oceans and Environment Branch Maritimes Region Bedford Institute of Oceanography P.O. Box 1006 Dartmouth, Nova Scotia B2Y 4A2 Direction des Océans et de L'environnement Région des Maritimes Institut océanographique de Bedford C.P. 1006 Dartmouth (Nouvelle-Écosse) B2Y 4A2 Compiled by V. E. Kostylev Compilé par V.E. Kostylev June 2002 / Juin 2002

Upload: vucong

Post on 25-Apr-2018

212 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

C S A SCanadian Science Advisory Secretariat

S C C SSecrétariat canadien de consultation scientifique

© Her Majesty the Queen in Right of Canada, 2002© Sa majesté la Reine, Chef du Canada, 2002

ISSN 1701-1280www.dfo-mpo.gc.ca/csas/

Proceedings Series 2002/023 Série des compte rendus 2002/023

Proceedings of a Benthic HabitatClassification Workshop

Compte rendu d’une Ateliersur l’habitat benthique

Meeting of the MaritimesRegional Advisory Process

Réunion du Processus consultatifdes provinces Maritimes

Maintenance of theDiversity of Ecosystem Types

Maintien de la diversitédes types d'écosystème

A Framework for the Conservation ofBenthic Communities of the Scotian-Fundy Area of the Maritimes Region

Cadre de conservation descommunautés benthiques du secteur

de Scotia-Fundy, Région des Maritimes

25 and 26 June 2001 Les 25 et 26 juin 2001Bedford Institute of Oceanography

Dartmouth, N.S.L’Institut océanographique de Bedford

Dartmouth, (N.-É.).

J. Arbour (Chair) J. Arbour (président)Oceans and Environment Branch

Maritimes RegionBedford Institute of Oceanography

P.O. Box 1006Dartmouth, Nova Scotia

B2Y 4A2

Direction des Océans et deL'environnement

Région des MaritimesInstitut océanographique de Bedford

C.P. 1006Dartmouth (Nouvelle-Écosse)

B2Y 4A2

Compiled by V. E. Kostylev Compilé par V.E. Kostylev

June 2002 / Juin 2002

Page 2: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

Proceedings of a Benthic Habitat Classification WorkshopMeeting of the Maritimes

Regional Advisory Process

Maintenance of theDiversity of Ecosystem Types

A Framework for the Conservation of Benthic Communitiesof the Scotian-Fundy Area of the Maritimes Region

25 and 26 June 2001Bedford Institute of Oceanography

Dartmouth, N.S.

J. Arbour (Chair)Oceans and Environment Branch

Maritimes RegionBedford Institute of Oceanography

P.O. Box 1006Dartmouth, Nova Scotia

B2Y 4A2

Compiled by V. E. Kostylev

June 2002

Page 3: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

2

Foreword Avant-propos

The purpose of this proceedings is to archivethe activities and discussions of the meeting,including research recommendations,uncertainties, and to provide a place toformally archive official minority opinions. Assuch, interpretations and opinions presentedin this report may be factually incorrect or mis-leading, but are included to record as faithfullyas possible what transpired at the meeting. Nostatements are to be taken as reflecting theconsensus of the meeting unless they areclearly identified as such. Moreover, additionalinformation and further review may result in achange of decision where tentative agreementhad been reached

Le présent compte rendu fait état desactivités et des discussions qui ont eu lieu àla réunion, notamment en ce qui concerneles recommandations de recherche et lesincertitudes; il sert aussi à consigner enbonne et due forme les opinions minoritairesofficielles. Les interprétations et opinions quiy sont présentées peuvent être incorrectessur le plan des faits ou trompeuses, maiselles sont intégrées au document pour quecelui-ci reflète le plus fidèlement possible cequi s’est dit à la réunion. Aucune déclarationne doit être considérée comme uneexpression du consensus des participants,sauf s’il est clairement indiqué qu’elle l’esteffectivement. En outre, des renseignementssupplémentaires et un plus ample examenpeuvent avoir pour effet de modifier unedécision qui avait fait l'objet d'un accordpréliminaire

Page 4: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

3

TABLE OF CONTENTS

Abstract / Résumé .............................................................................................................................. 4

Introduction ......................................................................................................................................... 7

Workshop Agenda ..............................................................................................................................10

Literature Review (B. Hatcher) ...........................................................................................................12

Summary of workshop Presentations.................................................................................................35

An Oceans Management Perspective on a Benthic Classification Framework (R. Rutherford) ........36

Objectives, Indicators, and Reference Points for Conserving Habitat (R. O’Boyle)...........................39

Availability of Information on Benthic Macroinvertebrate Communities for Classifying MarineEnvironments on the Scotian Shelf (P.L. Stewart and B. T. Hargrave)..............................................40

EUNIS Classification: An Excellent Example of a Well-developed General ClassificationSystem for Multiple Uses (P. Boudreau) ............................................................................................65

The Surficial Sediments of the Scotian Shelf: A Review of Published Maps, Limitations,Future Uses and Recent Advances is Seabed Mapping (G.B.J. Fader)............................................68

Challenges in Habitat Classification and Mapping (V.E. Kostylev) ....................................................72

Habitat Classification: A U.S. (NOAA Fisheries) Perspective (S.K. Brown).......................................77

Workshop Outputs..............................................................................................................................85

List of Attendees.................................................................................................................................93

Page 5: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

4

ABSTRACT RÉSUMÉ

In June 2000, a joint Fisheries and OceansCanada (DFO) Science-Oceans workshopwas held on ecosystem considerations forthe Eastern Scotian Shelf IntegratedManagement (ESSIM) Initiative. One of thekey outcomes of the workshop was thedecision that, based on present knowledge,the spatial distribution of benthic ecosystemtypes should be mapped within the ESSIMarea. It was agreed that a series of RegionalAdvisory Process (RAP) meetings should beheld to provide a framework and scientificbasis for ecosystem objectives concerningthe maintenance of benthic communities inthe Scotia-Fundy area of the MaritimesRegion. To complete the exercise, thefollowing phases were identified:

En juin 2000, les Directions des Sciences etdes Océans du ministères des Pêches et desOcéans (MPO) ont tenu ensemble un atelierau sujet des considérations relatives àl’écosystème dans le cadre du Programme degestion intégrée de l’est du plateau néo-écossais (ESSIM). Un des principaux résultatsde l’atelier a été la décision d’établir descartes, en fonction des connaissancesactuelles, de la distribution spatiale des typesd’écosystèmes benthiques dans la zone viséepar l’ESSIM. Il a été convenu de tenir unesérie de réunions dans le cadre du Processusconsultatif régional (PCR) afin d’établir lecadre et la base scientifique des objectifsécosystémiques concernant le maintien descommunautés benthiques dans le secteur deScotia-Fundy de la Région des Maritimes.Pour mener à bien ce projet, on s’est entendusur les phases suivantes :

• Phase I: Review existing approaches toclassifying and characterizing benthiccommunities, present and assess aproposed classification scheme, andrecommend a benthic classificationscheme for use in the Scotia-Fundy area.

• Phase I : Revoir les approches existantesen matière de classification et decaractérisation des communautésbenthiques, présenter et évaluer uneproposition de système de classification etrecommander un régime de classificationà adopter dans le secteur de Scotia-Fundy.

• Phase II: Apply the recommendedclassification scheme to the ScotianShelf to test and validate theclassification scheme, and produce amap of the benthic ecosystem.

• Phase II : Appliquer le système declassification recommandé au plateaunéo-écossais afin de le mettre à l’épreuveet de le valider, et produire une carte del’écosystème benthique.

• Phase III: Assess the sensitivity of theenvironments classified under thescheme to human activities prevalent onthe Scotian Shelf, and set referencepoints within DFO’s EcosystemObjectives Framework.

• Phase III : Évaluer la vulnérabilité desmilieux classés selon le système auxactivités anthropiques qui se déroulent surle plateau néo-écossais et établir despoints de référence au sein du cadre desobjectifs écosystémiques du MPO.

These proceedings provide the results of thePhase I workshop held on June 25th and 26th,2001. The workshop was attended by arange of experts including benthic scientists,ocean management specialists, industryrepresentatives and academia. Theobjectives and workshop structure were asfollows: (a) provide a review of existingapproaches used in benthic classification inCanada and internationally; (b) based on

Le présent compte rendu relate les résultatsde l’atelier de phase I, tenu les 25 et 26 juin2001. Des experts d’horizons divers ontassisté à cet atelier, dont des spécialistes desmilieux benthiques, des spécialistes de lagestion des océans, des représentants del’industrie et des universitaires. L’atelier visaitles objectifs suivants : a) revoir les approchesexistantes en matière de classificationbenthique au Canada et à l’échelle

Page 6: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

5

available information and present knowledge,present and review current classificationmaps of the benthic communities in the studyarea; and (c) recommend a classificationscheme for use in the Scotia-Fundy area ofthe Maritimes Region.

internationale; b) d’après l’information dont ondispose et les connaissances actuelles,présenter et examiner les cartes declassification courantes des communautésbenthiques dans la zone à l’étude et c)recommander un système de classification àutilier dans le secteur de Scotia-Fundy de laRégion des Maritimes.

Following the presentation of case studiesand research results by several participants,a plenary discussion session was held toprovide advice and direction for thedevelopment of a future benthic classificationscheme for the Scotia-Fundy area. Theconclusions of this workshop do notnecessarily represent the official policy or theviews of DFO. Efforts are currentlyunderway, as part of Phase II, to validate andtest the recommended scheme for theScotia-Fundy area.

Après la présentation d’études de cas et derésultats de recherches par plusieursparticipants, on a tenu une séance plénièredans le but de formuler un avis et uneorientation pour l’élaboration d’un futursystème de classification benthique pour lesecteur de Scotia-Fundy. Les conclusions decet atelier ne représentent pasnécessairement la politique officielle ou lepoint de vue du MPO. Dans le cadre de laphase II, on s’efforce actuellement de valideret mettre à l’épreuve le système recommandépour le secteur de Scotia-Fundy.

Page 7: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

(blank page)

Page 8: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

7

INTRODUCTION

The DFO Policy Committee tentatively defined seven objectives to guide ecosystem-basedmanagement. The first objective states that human activities should be managed so as tomaintain the diversity of ecosystem types within acceptable bounds. A science workshop washeld in June 2000 in support of the Eastern Scotian Shelf Integrated Management (ESSIM)initiative. It discussed indicators and reference points in relation to this objective but did not definespecifics, leaving this to future detailed study. Concerning benthic environments, the workshopconsidered that, based on present knowledge, the spatial distribution of benthic ecosystem typescould be mapped within the ESSIM area. Present ocean use activities could then besuperimposed on this map to determine the proportion of each type that may be disturbed. Thisanalysis would lead to the definition of areas in which there should be restricted or no activity. Itwas felt that the extent to which these measures would apply could vary by ecosystem type (e.g.deep corals may require a higher percentage of protection than sand – based communities).

A workshop to review the ecosystem objectives was held in Sidney, B.C., during 27 Feb –2 March. This workshop reaffirmed that the maintenance of diversity should occur at thecommunity, species and population level.

To further the objectives of ESSIM and related ocean management initiatives, the Maritimes JointBranch Management Committee (JBMC meeting of 11 September 2000) agreed that a RAPmeeting be held to provide the scientific basis for the maintenance of benthic communities in theScotian-Fundy Area of the Maritime Region.

Issues and Objectives

This meeting will provide an opportunity to review benthic communities and habitat diversity in theScotia-Fundy Region. It is recognized that some information from the pelagic zone and of fishdistribution patterns will need to be considered. The definition of communities for the planktonand fish assemblages will be deferred to a separate review and then the two will be integratedinto an ecosystem classification including all trophic levels and marine species.

There are three major phases to this exercise. The first will be to review existing approaches toclassifying/characterizing benthic communities, present and assess a proposed classificationscheme and recommend a benthic classification scheme for use in the Scotia Fundy area. Thesecond phase will involve the application of the classification scheme to the Scotian Shelf in orderto validate and test the scheme and to produce a map of the benthic ecosystem. The third phaseinvolves the assessment of the sensitivity of these environments classified under the system tohuman activities prevalent on the shelf and the setting of reference points within DFO’sEcosystem Objectives Framework.

The approach taken here will be to address these components through a series of separateworkshops. The first will address phase one and will provide the forum for developing therecommended benthic classification system. The second phase will be carried out throughcontract and in-house work and the third through a workshop that will be held at a later date toreview the application of that system in the Scotia Fundy area and the application of that schemeto the identification of sensitive habitats.

The geographic focus of the review will be the Scotia-Fundy area of the Maritime Region (4VWXplus 5 up to the US/Canada border). Given the additional complexities of data availability withinthe intertidal zone, the review will address information beyond the coastal fringe (i.e., deeper thanabout 10 meters in depth). The following are the specific objectives of the meetings.

Page 9: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

8

Phase 1:

Objective 1. Review the existing approaches used in Benthic Classification in Canada, the USand ICES.

Objective 2. Based on available information and present knowledge, present and review currentclassification maps of the benthic communities in the study area.

Objective 3. Provide a recommended classification scheme for use in the Scotia Fundy area ofthe Maritimes.

Phase 2:

Objective 4. Apply the classification and produce benthic ecosystem maps of the Scotian Shelf.

Objective 5. Validate the classification scheme and maps.

Phase 3:

Objective 6. Provide reference points that define the percentage area of each benthic communitythat needs to be protected by restrictions of uses that generate habitat impacts.

Objective 7. Document geographic patterns of human activity in the marine environment, and (tothe degree possible) provide a measure of the impact of each activity on the benthiccommunities defined under objective one.

Objective 8. Recommend next steps required for the selection of areas that could be zoned forrestricted use in order to achieve the putative conservation objective of"maintenance of benthic communities in the Scotia-Fundy area of the CanadianEEZ".

Products

Interim Products:

• Proceedings of the Phase 1 Workshop• Benthic Classification Model• Maps of the Benthic Ecosystems• Habitat Status Report (HSR) (2 reports)

1) recommended Benthic Classification System; and2) Sensitive Benthic Habitatsfor presentation to various advisory boards and general public.

Research Documents summarizing the technical basis for the conclusions and recommendationsin the HSR.

Proceedings document reporting the discussion of the RAP meeting.

Phase 2:

Application of the Classification Scheme

The classification system developed through workshop one will be applied to a pilot area on theScotian Shelf. Shortcomings in the model will be identified and the classifications systemmodified. The classification will then be applied to the Scotian Shelf and benthic ecosystem mapsproduced.

Page 10: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

9

Phase 3:

Assessment of Ecosystem Sensitivity

The third phase of the process will lead to an assessment of the sensitivity of the benthicecosystems identified in Phase 2 to human activities found or anticipated on the shelf. This will becarried out through a workshop which will focus on presenting the uses and activities on the shelf,an assessment of how these activities will impact on the benthic ecosystems of the shelf and thedevelopment of reference points for the benthic ecosystems relative to the disturbances expectedfrom these human activities.

Page 11: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

10

WORKSHOP AGENDA

Monday, 25 June 2001

09:00 – 09:10 Introduction (J. Arbour)

09:10 – 09:30 Workshop Objectives and Agenda (J. Arbour)The workshop background, objectives and agenda will be discussed. It will emphasized that weare looking for practical guidance and it is to be expected that as Science advances, we will needto change with it. What we will develop at the meeting will likely be applicable for 3 – 5 years andthen be updated. Also, the outline of the HSR will be given and discussed.

09:30 – 10:00 Ocean Management Needs (R. Rutherford)Use overview, issues, pressures and application of the information from this session. The use ofthe information produced by this RAP in the ocean management on the Scotian Shelf will bediscussed.

10:00 – 10:30 Break

10:30 – 11:00 Framework for setting Ecosystem Objectives (R. O’Boyle)A framework of objectives, indicators and reference points in relation to the maintenance ofbenthic communities will be presented. Particular attention will be given to the use of area as atool to manage human impacts. This will provide the context for the rest of the meeting’sdiscussion.

11:00 – 11:30 Data Availability (P. Stewart)An overview of the availability of data to support the description of the benthic communities will begiven.

11:30 – 12:00 A review of the Proposed ICES models for benthic classification (P. Boudreau)

12:00 – 13:00 Lunch

13:00 – 13:45 Geological Survey of Canada perspective (G. Fader and V. Kostylev)

13:45 – 14:15 A Definition of Benthic Diversity (E. Kenchington)An overview of benthic community diversity will be given, along with what characteristics need tobe measured. This will also aid in determination of classification needs.

14:15– 14:45 A review of the approach being taken by NOAA in the classification of marinehabitats. Classification systems in use in the US (S. Brown)

14:45 – 15:00 Break

15:00 – 17:00 Discussion Groups

Tuesday, 26 June 2001

09:00 -- 09:30 Review of Day 1 discussion groups

09:30 -- 10:30 A proposed classification system (B. Hatcher)

10:30 – 11:00 Break

Page 12: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

11

11:00 – 14:00 Working Groups (12:00-13:00 Lunch)Participants in the working groups will assess the proposed classification scheme, identify gapsand weaknesses, recommend ways of strengthening the scheme and propose a recommendedapproach.

14:00 – 15:30 Reporting Back from Groups

15:30 – 15:45 Break

15:45 – 16:30 PlenarySubsequent to the reports from the working group will be a facilitated discussion in plenary toidentify the agreed upon components for the classification system and resolve gaps anddifferences.

16:30 Next Steps(In the interim the recommended Classification Scheme will be applied to the Scotia FundyRegion and the associated maps produced.)

Page 13: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

12

LITERATURE REVIEW

B. HatcherCFCL-CBCL Ltd.

This review was prepared as contract deliverable for Department of Fisheries and Oceans(Canada).

The challenge facing the regional assessment process (RAP) is to derive a scientificallydefensible, legislatively robust and spatially-explicit benthic classification scheme for the massiveScotian shelf ecosystem. It must serve the needs of managers charged with the preservation ofbenthic biodiversity through the implementation of an ecosystem-based management policy forthis ocean management area (OMA). The goal of this document is to recommend a frameworkfor proceeding on the development of such a classification using the data, information, knowledgeand wisdom presently available. Elaboration on and disagreements with the points raised hereformed the basis of discussion at the RAP workshop, and may substantially alter the outcomes ofthe research as it progresses.

Objectives

• Present a discussion list of theoretical models and examples of ecological classification andzoning schemes that address the need for protecting benthic biodiversity.

• Assess the applicability of various benthic classification schemes to management planningfor the maintenance of ecosystem types on the Scotian Shelf, and document the justificationfor the use of a range of ecological indicators and reference points. Identify gaps in scientificknowledge.

• Evaluate the benthic ecosystem types on the Scotian Shelf and justify proposed guidelinesfor conservation objectives for each type.

The Need to Maintain a Diversity of Marine Ecosystem Types on the Scotian Shelf

The reality and urgency of the need is taken as a given from the terms of reference for thisresearch, and will not be reiterated or elaborated here. Simply put, the essential goal ofecosystem approaches to management of the Scotian Shelf ocean management area is topreserve its biodiversity, and thereby sustain the ecosystem goods and services provided to thepeople of Atlantic Canada, and to the rest of the world. The challenge is to provide robustdecision-support for precautionary and adaptive management in the short-to-medium term. Theprecautionary principle requires that steps be taken to preserve the Scotian Shelf’s biodiversity inthe absence of complete knowledge of its magnitude or distribution, or of the anthropogeniceffects on it. A benthic habitat classification scheme, albeit incomplete at the species level and atall spatio-temporal scales, is an essential part of the framework for action by the Department ofFisheries and Oceans.

Methods

A library and internet literature review was conducted using the following key words:Marine continental shelf ecosystem(s), Marine continental slope ecosystem(s), Marine habitatclassification, Marine habitat mapping, Seabed mapping, Marine benthic ecosystem structure,Marine benthic ecosystem dynamics, Benthic habitat classification, Benthic habitat mapping,Benthic zonation, Benthic community structure, Benthic community organization, Spatialdistribution of benthos, Critical marine habitat, Marine biodiversity conservation.

A comprehensive table of all known marine ecosystem classification criteria was drawn up fromthe information presented in these references, and circulated to project staff on 22 January 2001.

Page 14: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

13

Interviews were held with twenty-four (24) DFO scientists holding direct experience and data onScotian Shelf ecosystems (including the Bay of Fundy). The information thus obtained was usedto discover new concepts, research results and literature, fill in the classification table, and seekadvice on priorities and progress in ecosystem classification of the Scotian Shelf. A discussion listof theoretical models and examples was circulated to project staff on 15 February 2001.

Based on these inputs, the characteristic metrics, domains, scales, grains and densities of datarelevant to each criteria are being compiled for the Scotian Shelf (ongoing).

A draft research discussion document was circulated to project staff on 31 March 2001. A finaldiscussion document, revised in light of feedback to the draft was circulated on the first day of theRAP meeting on 25 June 2001.

Definition of Terms

The following definitions are adopted for the purposes of this study. They represent common,current usage, but are not meant to be exclusive. Many analogous terms that are unique tospecific classification schemes (e.g. Wilken, et al., 1996) or are not in common use are notexplicitly defined here. It is crucial that the terms be clearly defined and consistently used in thecontext of this research. Elaborations or disagreements with these will form the basis ofdiscussion, and may substantially alter the outcomes of the research.

Scotian Shelf extends from the SW edge of the Laurentian Channel to the NE edge of the greatNortheast Channel and its extension to the US border in the Gulf of Maine; from the landwardmargin of the sublittoral zone (approximately 10m depth) to the shelfward edge of the abyssalzone (approximately 1200m depth). It thus includes the SE continental slope and the Bay ofFundy, and excludes the Georges Bank. It is a complex section of a NW Atlantic large marineecosystem (LME) that includes all of the neritic and the continental shelf margin of the oceanicprovince. The area has been designated by the DFO as ocean management area (OMA).

Environment is the inorganic, non-living surroundings in which organisms, populations,assemblages and communities exist. The spatial scales of the environment are determined by itsphysical structure and the physical forcing. (Thus, we speak of the “tidally-dominated benthicenvironment” or the “deep slope environment”.)

Habitat is the characteristic environment in which a particular organism, population, assemblageand community lives. It includes the living and non-living organic components that modify therelevant environment. The spatial scales of habitats are defined by the interaction between theenvironment and the biota of concern, and may thus be characterized by the non-living or livingattributes. (Thus, we speak of the “cobble-silt habitat” and also of “the kelp bed habitat” or “lobsterhabitat”.)

Community is a characteristic assemblage of organisms that has repeatable and predictablestructure, function, habitat-association and environmental affinities. The spatial scales ofcommunities are defined by the interaction between the physical-chemical boundaries of theirhabitats and the overlapping distributions of the community components. Operationally, we firstdescribe an assemblage of organisms in a definable habitat and then, as we learn more about theecological interactions and dynamics of the component populations, we may come to recognize acommunity of organisms with characteristic attributes. (Thus, we speak of the benthic community,and also of the kelp bed community.)

Assemblage is a group of organisms that occurs within an environment (no implication ofrepeatable structure or predictable function). The spatial scales of assemblages are defined bythe extent of the sampling regime and the habitat they occupy. (Thus, we speak of the filterfeeder assemblage and the deep coral assemblage.)

Page 15: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

14

Population is an interacting group of organisms of the same species. Populations are furtherclassified as local, meta and species populations according to the intensity of interaction amongstindividuals and spatially separated groups within them (“stocks” may be defined at any of theselevels depending on the relationship between the fishery and the species and its environment).The spatial scales of populations are defined by their distributions, generally increasing from localto meta to global species populations.

Benthos is the sum total of life forms in the benthic and epibenthic habitats: organisms,populations, assemblages and communities. As such, the benthos spans the entire range ofspatial scales of the shelf.

Substratum is the non-living matrix in, on and immediately over which infaunal, benthic andepibenthic organisms, assemblages and communities exist (i.e. it is the solid component of theenvironment). Attributes of the marine benthic substrata include (at increasing spatial scales):sediment grain size, sediment porosity, sediment thickness, outcrop hardness-friability, rugosity,topographic relief, bathymetry and complexity (described using e.g. fractal dimension).

Ecosystem is an interacting assemblage of organisms and their environment in which the sum ofthe internal transfers of biogenic materials or energy exceeds the transboundary fluxes. Thetypology of ecosystems includes aspects of both the non-living and living components.Ecosystems are usually defined by the environment (e.g. the “shelf-edge canyon ecosystem”),often defined by the dominant forcing (e.g. the “upwelling ecosystem”), and sometimes defined bythe characteristic community within them (e.g. the “coral reef”). Ecosystems may exist across thefull range of spatial scales, from an interstitial space, to a large marine ecosystem such as theGulf of Maine, to the ecosphere. The definition is first operational, and the spatial domain follows.(A trite, but not unrealistic alternative definition is that an ecosystem is what a competentecologist says it is!)

Ecological Unit is the lowest level in a hierarchical classification of ecosystems. Units arecontiguous and not repeatable (i.e. each is unique), but there may be more than one of the typeor next higher level of classification. Thus, we may speak of “fields of sand waves” that occur inplaces where near-bed tidal flows over unconsolidated sediments are very strong, but the field ofsand waves at the SW entrance to Halifax’s outer harbour is a unique ecological unit - there isonly one.

Biodiversity is the variety of life forms inhabiting a defined area. As the spatial scale increases, itincludes progressively more levels of the hierarchies of biological and ecological organization (i.e.ecosystems): genes, genotypes, phenotypes, organisms, populations, assemblages,communities, habitats, biotopes, etc. Emergent properties at the levels of organization above thespecies population mean that the whole is greater than the sum of the parts, and must bepreserved as such if the ecosystem functions inherent to a biodiverse structure are to bemaintained.

Classification is the process by which mapable units of biological organization are identified. Aclassification scheme provides the framework of criteria upon which biological units are assignedto a spatially-distinct group. (Note well the distinction from the taxonomic classification of benthos,which is space-independent.)

The Requirement for a Marine Habitat Classification Scheme

The spatial extent and the diversity of known marine habitats of the Scotian Shelf precludescomplete knowledge of the distribution and abundance of all the life forms that contribute tomarine biodiversity: generalizations about the distribution of these life forms must be made athigher taxonomic levels than the individual or species. That life forms are hierarchicallydistributed across multiple level of organization in predictable patterns is a “given” of ecology. Therules of assembly for these units in any given environment are poorly understood however, andoperate at multiple scales of space and time: generalizations can only be made for the units that

Page 16: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

15

persist at larger, longer scales (i.e. habitats). The variety of (often overlapping) human uses of theScotian Shelf OMA demands spatially explicit management (i.e. zoning) that must be informed bythe distribution of habitats. It follows that a marine habitat classification is an absolute pre-requisite for any management strategy designed to maintain a diversity of ecosystem types at thescales of the Scotian Shelf. The focus is first on a classification of benthic habitats because thespatial stability of seabed compared to the water masses makes it easier to map and zone. Themajor contribution of benthos to marine biodiversity, and the demonstrated impacts of manyhuman activities on benthic communities also argue that priority be given to the classification ofbenthic habitats.

Ecological Classification and Zoning Schemes that Address the Need for ProtectingBenthic Biodiversity on the Scotian Shelf

The requirement for a geo-referenced classification scheme for marine and coastal ecosystemsto support management policy and implementation based on the zoning of human activities isnear-universally accepted in the DFO, in Canada and in most other maritime nations. In thecontext of DFO’s mandate for the Scotian Shelf, the scheme must identify indicators of benthicbiodiversity and reference points against which management performance can be assessed. Thisrequirement for defensible regulatory support, as well as the operational limits on monitoring,control and surveillance of ocean space at the scale of the Scotian shelf influence the goals andscales of the habitat classification scheme that might be employed.

There is as yet no consensus on the structure of that classification. This is in part because thereis recognition that different classifications are required for different applications (e.g. thepreservation of rare and endangered species at risk vs. the maintenance of the diversity ofecosystem types), and in part because the theoretical underpinnings of such classifications arepoorly developed.

Several organizations and agencies throughout the world are researching and using classificationschemes for marine benthic ecosystems. Many scientists and managers within the DFO andseveral other relevant agencies in eastern Canada are actively engaged in developingclassifications and maps of the marine ecosystems of the Scotian Shelf (including the Bay ofFundy).

Even recognizing the importance of the benthos to marine biodiversity, there are strongarguments for including water column attributes and processes in a list of criteria for benthicclassification. These include the importance to benthic community structure and function ofsedimented organics derived from water column production, and the magnitudes of benthic-pelagic coupling in energy, nutrient and reproductive fluxes.

There are strong arguments for including littoral and coastal attributes in a list of criteria forbenthic classification of certain parts of the Scotian Shelf (i.e. the Bay of Fundy). These includethe importance to benthic community structure and function of terrestrial inputs of organics andnutrients, and intertidal resuspension and transport of materials and organisms.

The size of the Scotian Shelf domain means that it will never be fully sampled. Therefore, anybenthic classification scheme that supports zoning for biodiversity preservation must involvescaling up from sample areas to the shelf scale. The selection of a suitable framework forclassification is largely predicated on its ability to up-scale to that of the ocean management areawith sufficient spatial accuracy for the management actions intended (e.g. multiple use zoning,MPAs, etc.).

Temporal variation in the abundance and spatial distribution of benthic organisms is manifold atmany scales. Any map of benthic communities will be a snapshot. The classification frameworkmust optimize the match of temporal-spatial scales such that the zonation depicted has ecologicaland practical meaning over typical management decision periods.

Page 17: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

16

There are compelling arguments for using a hierarchical classification system for the Scotianshelf’s ecosystems. The processes which structure benthic communities, the data which areavailable to describe pattern, and the purposes to which the mapped outputs will be applied allextend across a range of organizational levels and scales of space and time.

Types of Benthic Classification Schemes:

Spatial classifications of marine benthos are based on process (ecological function) and pattern(ecological structure). The approaches are not mutually exclusive, as structure and function areintimately linked at all levels of biological organization. Most schemes incorporate both aspects,but process-based classifications focus either at the very large (macro) scales of biogeographywhere geological, oceanographic and evolutionary processes dominate, or at the very small(micro) scales of settlement and competition where biological processes are important. Formapping purposes, process-based classifications require an underlying, theoretical basis (model)that relates the process to the distribution of biotic units. Pattern-based classifications span thefull range of spatial scales, but are dependent upon (and limited by) empirical descriptions of thedistribution of biotic units. Process-based classifications may be developed in the absence ofempirical data (although such are essential for calibration and verification), and thus are morelikely to result from deductive approaches. Pattern based classifications may be developed aposteriori, in the absence of a theoretical model, and are thus often derived by induction.

The development and application of landscape (i.e. terrestrial) classifications and habitat mapshave a longer history than those of seascapes. The empirical tools are particularly welldeveloped, perhaps because of the relative ease of sampling the land, and the analytical toolsare directly applicable to the marine realm. Landscape ecology has fewer lessons for process-based marine classification because of the fundamental differences between the ocean and landenvironments and between the adaptations of marine and terrestrial organisms.

For practical reasons, hierarchical classifications usually use the distribution of major physicalstructures and processes to map benthos at the higher levels of organization and larger spatialscales (i.e. communities and above at shelf scales), and the distribution of biota to map benthosat the community level and below. This reflects both the scale of process that determine patternin the benthos, and the density of empirical data available. New technologies of remote sensingand underwater visualization are releasing these constraints.

A major challenge is to identify an adequate (and ultimately an optimal) scheme of classificationfor protecting marine benthic biodiversity. The fact that biodiversity is hierarchical argues ahierarchical approach to classification and mapping. The current level of human use of andimpact on marine environments and biota, the precautionary principle, and the perceived urgencyin statute and soft law as well as some public opinion recognize the necessity of a defensiblebenthic classification scheme to support adaptive management decisions in the near future.Expediency requires that the distinction be made between scientific models designed to improveunderstanding of how nature works, and those designed to support the management of humanbehaviour. There is also a necessity to match the scales of habitat mapping to the scales ofhuman activities.

There are at least 40 classification systems that have been developed to some degree for marineecosystems (but less than 20 have been applied). It is appropriate to synthesize the strengthsand weaknesses of these in the context of the goal for the Scotian Shelf. The following survey oftheoretical models and examples of benthic classification and zoning schemes is not meant to beexhaustive, as the exercise has been undertaken recently by several authors (e.g. Frith, et al.,1993; Robinson and Levings, 1995; Watson, 1997; Buchanan and Christian, 1999; NCER, 2000).The results represent priorities for further research, based on the information and wisdomgathered to date.

Biogeographic zonation schemes use empirical knowledge of the geographic extent and spatialdistribution of benthos units that are stable at ecological to evolutionary time scales. The typical

Page 18: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

17

biological metric is the presence-absence or density of benthic organisms. The distributions arederived from published records of species occurrence and from interpolations among benthicgrab, photo or video samples. A posteriori, multivariate classification or ordination techniques areoften used to identify populations of indicator species, characteristic assemblages of associatedorganisms or distinct benthic communities. The analysis of biota belonging to many taxa sampledover many years and spatial scales leads to the identification of biogeographic provinces atcontinental shelf scales, while more localized and denser sampling is used to produce maps ofecological units at habitat scales.

The ecological processes that control the associations between biota and geographic variablessuch as latitude and depth are inferred, and need not be explicit or understood for the applicationof the scheme. None-the-less, spatial correlations with environmental variables such astemperature and circulation cells are often assumed to be causal relationships.

Examples of Biogeographic zonation schemes include:- CSIRO’s (1996) Interim Marine Bioregionalisation for Australia- Sullivan Sealy and Bustamente’s (1999) EcoRegions- NOAA’s Benthic Habitat Classification Scheme (Coyne, et al., 2001)- Roberts and Hawkin’s SeaMap http://www.ncl.ac.uk/seamap (accessed 27 May 2001)- Mahon, et al.’s (1998) east coast NA demersal fish assemblages- Gomes, et al.’s (1992) Grand Banks demersal fish assemblages

Oceanographic control schemes apply theoretical or empirical relationships between benthos andhydrodynamic-hydrologic attributes of water masses to known distributions of physicaloceanographic features. Typical hydrodynamic metrics are exchange turnover rate, residencetime, stratification, mixing, current velocity and bottom shear velocity. Typical hydrologic metricsare temperature, salinity, and their temporal variance. Their spatial distribution may bedetermined from interpolations among hydrographic stations (e.g. T-S fields), mapped withremote sensing tools (e.g. Synthetic Aperature Radar), or predicted from numerical models (e.g.CANDIE). Biota are related to these attributes in terms of presence-absence and sometimesabundance or biomass. The typical biological unit is the species, assemblage or community. Theassociations may be determined a priori from the literature (e.g. research on the effect of near-bed current on feeding and dispersion of macro-invertebrates), or a posteriori from classificationanalysis of biota sampled under different hydrological regimes). The ecological processes thatdetermine the associations between biota and water masses are usually explicit and wellunderstood.

Examples of Hydrodynamic-Hydrological control schemes:- Marine Environmental Quality Advisory Group’s (MEQAG, 1994)- Longhurst’s (2001) Marine production zones- Loder, et al.’s (2001) Scotia-Maine hydrographic and circulation regimes- LOICZ’s Typology of coastal and marine environments

http://water.kgs.ukans.edu:888/public/Typpages/index/html (accessed 17 June 2001)- Wolanski’s (2000) Great Barrier Reef oceanographic zones

Geomorphological correlation schemes apply empirically known associations between benthosand substratum type to known or extrapolated distributions of substratum. Typical substratummetrics are seabed relief, rugosity, sediment thickness and grain size. Their spatial distribution iseither interpolated from point photo or grab samples (geological maps), or mapped with remotesensing tools (e.g. side-scan sonar). Biota are related to these attributes in terms of presence-absence and sometimes abundance or biomass. The typical biological unit is the species or theassemblage. The associations may be determined a priori from the literature (e.g. research onthe abilities of infauna to survive in different types of sediment), or a posteriori from classificationanalysis of biota sampled on different substrata. The ecological processes that control theassociations between biota and substratum are inferred, and need not be explicit, or evenunderstood for the application of the scheme.

Page 19: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

18

Examples of Geomorphological correlation schemes:- Andrefouet, et al.’s (2001) Typology of atolls- Davis et al.’s (1994) Biophysical classification of offshore regions of N.S.- Day and Roff’s (2000) Framework classification of Canada’s oceans- GBRMPA’s Great Barrier Reef marine ecoregions.

http://www.gbrmpa.gov.au/corp_site/key_issues/conservation/rep_areas/documents/bioregions_2001_06.pdf (accessed 13 March 2001)

- Kostylev, et al.‘s (2001) Benthic habitat mapping on the Scotian Shelf- Mumby and Harborne’s (1999) Caribbean marine habitat classification scheme

Ecological control schemes use established models of biological processes, interactions andfluxes of biomaterials to predict the distributions of organisms, populations and communities.

Typical metrics include genetic distance, recruitment rate, primary production, ecotrophicefficiency, organic sedimentation rate, etc. These ecological inputs to the models may bemeasured directly, derived from linked bio-physical and biogeochemical models, or assumed fromliterature values.

Model outputs indicate the presence and density of benthos as a function of the ecological controlprocess and input parameter values. Determining the spatial distribution of ecological unitsrequires that the models be spatially explicit, or that the distribution of the input parameters beknown (e.g. the spatial pattern of primary production).

Limitations of models and data mean that these classification schemes are usually locale-specificand small in scale.

Examples of Ecological control schemes:- Bruno and Bertness’s (2001) benthic habitat facilitation model- Massell and Done’s (1993) hydrodynamic classification of wave effects on coral benthos- Minn’s (1997) fish habitat identifications- Sousa’s (2001) disturbance models

Hybrid geo-physical or bio-physical schemes use a combination of the physical and biologicalschemes outlined above.

Typically, the spatial distribution of physical environments is further partitioned by known affinitiesor occurrences of characteristic biota.

Examples of hybrid classification schemes:- Harper, et al.‘s (1983) Marine regions of Canada- Hiscock’s (1995) and Connor, et al.‘s (1997) Classification of benthic marine biotopes in the

U.K.- Australian Environmental Conservation Council’s (EEC, 2000) Marine habitat classes- Frith, et al.‘s (1993) Habitat classification systems for coastal B.C.- Hooper’s (1997) Coastal marine habitat classification system for Nfld.- Deither’s (1990) Marine and estuarine classification system for Washington Greene, et al.’s

(1998) Classification of deep seafloor habitats- Brown’s (1993) Classification system of marine and estuarine habitats

Hierarchical Hybrid schemes use a combination of physical and biological schemes arranged inorganizational and spatial hierarchies such that the scheme used for any given level provides themost powerful prediction of the distribution of biota.

Examples of Hierarchical Hybrid schemes considered:- Cowardin, et al.‘s (1970) Classification of wetlands and deepwater habitats of the United

States- Environment Canada’s Marine Ecological Classification System (MEQAG, 1994)

Page 20: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

19

- Zacharias, et al.’s (1998) British Columbia marine ecosystem classification- DFO Coastal and Estuarine classification for fish habitat (Williams, 1989)- European EUNIS marine habitat classification (Davies and Moss, 1999)

http://mrw.wallonie.be/dgrne/sibw/EUNIS/eunis.fulllistA.html (accessed 8 June 2001)- Hatcher, et al.‘s (1990) Marine habitat classification of the Abrolhos

From this categorized list it may be seen that benthic classification schemes developed to dateuse a wide range of criteria, may be categorical or hierarchical, and non-dimensional or explicitlyspatial. It is apparent that different schemes are developed to meet different needs, and not allschemes need, or result in actual maps of the spatial distribution of classes. Spatialclassifications of marine benthos are based primarily on pattern (ecological structure), while non-dimensional classifications depend more on process (ecological function). The approaches arenot mutually exclusive, as structure and function are intimately linked at all levels of biologicalorganization. Most schemes incorporate both aspects, but process-based classifications focuseither at the very large (macro) scales of biogeography where geological, oceanographic andevolutionary processes dominate, or at the very small (micro) scales of settlement andcompetition where biological processes are important.

For mapping purposes, process-based classifications require an underlying, theoretical basis(model) that relates the process to the distribution of biotic units. Pattern-based classificationsspan the full range of spatial scales, but are dependent upon (and limited by) empiricaldescriptions of the distribution of biotic units. Process-based classifications may be developed inthe absence of empirical data (although such are essential for calibration and verification).Pattern based classifications may be developed a posteriori, and may proceed in the absence ofa theoretical model.

The development and application of landscape (i.e. terrestrial) classifications and habitat mapshave a longer history than those of seascapes. The empirical tools (e.g. vegetation mapping) areparticularly well developed, in part because of the relative ease of sampling the terrestrialenvironment, and the analytical tools (e.g. GIS) are directly applicable to the marine realm.Landscape ecology has fewer lessons for process-based marine classification because of thefundamental differences between the ocean and land environments and between the adaptationsof marine and terrestrial organisms.

For practical reasons, hierarchical classifications usually use the distribution of major physicalstructures and processes to map benthic habitats at the higher levels of organisation and largerspatial scales (i.e. communities and above at shelf scales), and the distribution of biota to mapbenthos at the community level and below. This reflects both the scale of process that determinepattern in the benthos, and the density of empirical data available. New technologies of remotesensing and underwater visualisation are releasing these constraints.

A major challenge is to identify an adequate (and ultimately an optimal) scheme of classificationfor protecting marine benthic biodiversity. The fact that biodiversity is hierarchical argues ahierarchical approach to classification and mapping. The current level of human use of andimpact on marine environments and biota, the precautionary principle, and the perceived urgencyin statute and soft law as well as some public opinion highlight the necessity for a defensible,spatially explicit benthic classification scheme to support ocean planning and zoning decisions inthe near future. Expediency requires that the distinction be made between scientific classificationmodels designed to improve understanding of how nature works, and those designed to supportthe management of human behaviour. There is also a necessity to match the scales of habitatmapping to the scales of human activities in the ocean, and to the scale at which managementcan be effective (i.e. the area of the minimum manageable unit that can be monitored andcontrolled, typically 10km2 near shore and 100km2 offshore).

Page 21: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

20

Generalized Applications of Habitat Classification to the Scotian Shelf

All classification schemes of marine benthic communities developed for management decisionsupport (see recent reviews by Watson,1997; Buchanan and Christian, 1999; NCER, 2000; aswell as overview in previous section) are intended to result in quantitatively mapable units (i.e.they are spatially explicit models that seek to define how much of what will be found where). Themapable unit selected in these schemes is more often than not the habitat . This choice also hassome support in the theoretical literature. Southwood (1977) calls habitats “the templates forecology”, upon which rest all ecosystem components. Gray (1997) defines marine habitats asunits with distinct physical and biotic features, and goes on to note that their fragmentation is thegreatest threat to biodiversity (a conclusion also supported by the GESAMP, 2001). I believe thatthis analysis of threat is one-dimensional, and suggest, conversely, that the homogenization ofbenthic marine habitats is the greatest threat to marine biodiversity. Bottom dragging and trawlingare but one of several human activities that smear and break down habitat boundaries andstructures in the ocean environment, with increasingly well-known negative effects on biodiversity(Auster, 1998).

Whatever the classification criteria selection or the unit selected for mapping, there appear to betwo basic approaches to achieving the goal of defining how much of what is where: the deductiveand the inductive.

The deductive approach classifies benthic communities a priori from analysis (often mapping) ofthe physical environment. The method uses established or assumed ecological classificationmodels to infer the types and spatial pattern of benthos from first principles, and from knowledgeof the nature, magnitude and distribution of forcing or structuring attributes of the physical-chemical environment. Because of its dependence upon physical attributes, this might be calledthe “physical” mapping approach to marine benthic classification.

The inductive approach classifies benthic communities a posteriori from analysis of theassociations among biota. The method uses pattern searching and defining tools (usuallystatistical) to infer the spatial pattern and classification and of benthos from direct or remotesampling of benthic biota. Because of its dependence upon biological attributes, this might becalled the “biological” mapping approach to marine benthic classification.

The deductive (i.e. physical) approach is by far the most common for areas of the size of theScotian Shelf because of the impossibility of sampling the entire domain at a spatial densityadequate to resolve the distribution and abundance of benthic organisms in the finer grainhabitats. It also avoids the major problem of temporally variable distributions of organisms, whichrequires periodic, long term monitoring to resolve. The models are typically (but not always)hierarchical, derived from terrestrial (landscape) ecology, and based on known or assumedrelationships between physical-chemical attributes of the environment and characteristicassemblages of macroflora and macrofauna (Table 1). The mapping aspects are almost alwayshandled using remote sensing technologies, and increasingly using geographical informationsystems (e.g. Fader, et al., 2000; CEIMATE, 2000), and it is the ability of these tools to producesynoptic maps of vast areas that is the great strength of this approach to benthic habitatclassification.

Limitations of the physical-chemical data sets, the accuracy of their measurement and thetechnologies for mapping them constrain the power of this approach, but the real challenge (andcertainly the art) lies in the selection of physical classification parameters to be measured andmapped. The classification of biological communities by this method is only as good as therobustness and fidelity of the assumed relationships between the physical environment and thebiota. When insightful choices of parameters are made: these relationships can be statisticallyrobust (e.g. Kostylev, et al., 2001), and may be perfectly adequate for classifying benthiccommunities at the spatial scale of minimum manageable unit (MMU). The risk lies in committingto expensive programmes of measurement and mapping of physical parameters that do notpredict the distribution and abundance of benthic communities. This risk is inversely scale-

Page 22: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

21

dependent, being lowest at shelf scales (i.e. 100 to 1000 km) where physical controls dominatethe coarse grained distribution and abundance of biota, and greatest at sub-habitat scales (i.e.0.01 to 10 km) where biological controls may dominate (Levin, 1992). The MMU for the ScotianShelf (approx. 10km x 10km) clearly falls in between these two spatial domains, and hence therisk is real, but probably manageable.

Davis and Browne (1994) and Davis, et al., (1994) provide an example of the deductive approachby building on the hierarchical, “ecozone” classification of Canada’s marine regions developed forNational Parks by Harper, et al., (1983; 1993) and Hiroven, et al., (1995). These are essentiallygeographical classifications, stated by the authors to be based on geology and physiography,rather than oceanography. A wide range of classification parameters (including thoseoceanographic) are apparently used, however, albeit not in a quantitative fashion. Four (4)“districts” corresponding to across shelf location (i.e. inner, middle, outer and slope) arerecognized, and eleven (11) classes of “units’ that include nearshore water bodies such as baysand lakes, and offshore features such as bank tops, saddles, basins, plains, valleys, channelsand canyons. Each unit is further sub-divided at the next level of the hierarchy into pelagic andbenthic “habitats”, yielding a total of 58 unique habitats (the later ranging in size from approx.4,800 to 48,000 km2). The classification does not distinguish between the boundaries of benthicand pelagic habitats (a patently unrealistic assumption for areas that are not hydrodynamicallyconstrained by topography). While characterizations of typical sediment, hydrological and bioticattributes of each unit are made, no attempt is made to identify benthic communities from thismapping exercise.

Day and Roff (2000) provide another example of the deductive approach to classifying allCanadian marine space, with a tentative application to the Scotian Shelf. Seven (7) physicalparameters belonging to four (4) hierarchical levels are used to delineate nine (9) “naturalregions” and 62 unique “seascape” units (the later ranging in size from approximately 30 to370,000 km2). No attempt is made to identify benthic communities from this mapping exercise,however.

The inductive (i.e. biological) approach is more common for smaller, and usually shallower areaswhere dense sampling and direct observation of the benthos are feasible. The approach drawson a large body of ecological and biogeographical knowledge and practice in the sampling andorganisms to describe community structure and function, ranging in scale from tide pools toocean basins (e.g. Hedgepeth, 1957; Levinton, 1982; Polis, et al., 2001). Information may beobtained by actually collecting organisms (using SCUBA or with trawls and grabs), or byinspection and enumeration by SCUBA divers or by viewing U/W photos or video images. Thederivation of spatial pattern from such data draws on an equally strong body of knowledge andtools for analysing multivariate data (e.g. Pielou, 1972; Andrew and Mapstone, 1987; Warwickand Clarke, 1998). The great strength of this approach is that it actually measures the diversity,abundance and distribution of organisms within the sample area: the dependence onenvironmental surrogates (i.e. correlates) is removed. The limitations of the approach areessentially those of sampling offshore marine ecosystems: logistic challenges of work at sea andvast, unseen areas to sample, equivocal criteria for sample stratification, and highly variable ratesof temporal change in ecological processes affecting community structure. To this problem isadded the inherent unpredictability of community structure at the level of the local speciespopulation.

Most of the well-developed examples of the inductive approach come from Europe, where aseries of marine habitat characterization and mapping initiatives (e.g. Earll, 1992; Hiscock andConner, 1995; Conner, et al., 1995; Nijkamp and Peet, 1994) are coalescing towards a uniformclassification scheme based on characteristic assemblages of marine organisms (e.g. BioMar,MarLIN, EUNIS). Similar initiatives are underway in Australia (Lyne and Last, 2001). Theseactivities have focused primarily on shallow water environments, and draw on a series ofextensive, long-term investments in marine sampling. In these schemes, the geological andoceanographic parameters are associated with benthic communities a posteriori, and not, to date,in a quantitative fashion.

Page 23: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

22

Stewart, et al., (2001) have recently compiled data from 122 benthic grab sample stations on theScotian Shelf, and a further 310 in the bay of Fundy. Hargrave subsequently related thesequalitatively to the geo-physical habitat classification of Davis and Brown (1994), recognizing theinner shelf, banks, basins and slope as the primary zones, and substrate type in terms ofsediment grain size (4 ranges) and proportion of hard substratum (3 ranges) as the next lowerlevel of classification.

Hargrave (unpub.) concludes that sedimentary facies determine the benthic community type (asdefined by the major taxa of macrofauna), while diversity (in terms of species composition) isdetermined by habitat complexity, predation and other biological factors. He further suggests thatbenthic, macrofaunal biomass is determined by the food supply, along a gradient of diminishingsupply across-shelf. Hargrave’s conclusions are in broad agreement with the literature on factorscontrolling benthic community structure in deep sedimentary environments, but the lack of explicitrecognition of historical factors relating to disturbance and succession ignores a large body ofunderstanding derived from experimental work in shallow, hard bottom communities. They mayexplain why relationships between metabolism or turnover and position along the cross-shelf (i.e.food supply) gradient are not apparent. The major limitation on this work, however, is the smallarea sampled (<1m2 per station) and low density of sample stations (ranging from approx. 90 per12,350km2 in the Bay of Fundy, to <50 on the Western shelf, to <1 on the remainder of theScotian Shelf. For the purposes of defining indicators and setting reference points for benthicbiodiversity preservation, the problem of inadequate sampling rules out a strictly inductiveapproach.

Kostylev, et al., (2001) have combined the two approaches for the Browns bank area of theScotian shelf, and more recently, in the area of the Gully on the Eastern Shelf. Multibeam remotesensing, augmented by side-scan sonar and seismic profiles was ground-truthed with grabsamples and still photography to sample the megabenthos at 24 and 115 stations respectively.Seven (7) benthic habitats are recognized a posteriori on the basis of cluster analyses. Theidentified ‘habitats’ are actually a mixture of bio-physical attributes that may be considered asbenthic community types. Two habitats are simply charaterized by depth (shallow vs. deep), oneby broad trophic category (deposit feeders), and three by dominant genera or species). Thedistribution of the communities is related to four (4) sediment types, four (4) depth zones, three(3) relative estimates of near-bed current speed, and two (2) measures of temperature: annualmean and seasonal variance. Three (3) relative levels of a derived factor: substrate structuralcomplexity (related to grain size), are added to the classification but not rigorously assessed.Parametric statistical tests showed highly significant differences between these habitat types dueto the sediment type factor, while the co-variate depth did not exert a significant effect. Muti-variate tests distinguished two factors that explained about 30% of the observed variance in thedistribution of megafauna taxa, and about 42% of the observed variability in abundance.Sediment type was by far the most important correlate in these analyses, while depth-relatedfactors of the hydrodynamic and thermal regime were significantly correlated, but explained thesmaller proportions of the partitioned variance. These power of these relationships is similar tothat obtained in other studies of this sort at these scales (e.g. Warwick and Uncles, 1980;Warwick, 1988; Clarke and Ainsworth, 1993). At only 30 to 40% of the variance explained, theytend to be at the lower end of the range of 30 to 90% of variance in macrofaunal distribution andabundance that can be assigned to physical variables in intertidal and pelagic environments(Mann and Lazier, 1996; Lengendre and Legendre, 1998; Denny and Wethey, 2001).

An important question is whether the spatial accuracy and repeatability of these classes ofbenthic community type can be substantially improved. If so, whether most cost-effectively byincreasing the number or range of physical-chemical parameters in the analysis; increasing theaccuracy of measurement or estimation of the variates already used, or by increasing the densityand detail of the benthos sampling. At the scales of the MMUs of the Scotian shelf (typically100km2), and given the physical-chemical data sets already available for the area, the firstoptions is certainly a priority.

Page 24: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

23

A clear gap in virtually all the analyses undertaken to date is the lack of explicit inclusion ofmeasures of community function in the classification schemes. The two ecological processes bestknown to influence benthic community structure are the frequency and intensity of physicaldisturbance, and the rate and form of delivery of consumable organic material. Both of thesefactors can be directly and quantitatively linked to species diversity (e.g. Connell, 1978; Tyler,1986). Both factors are, in turn, determined by the oceanographic, hydrodynamic, and, in somecases, fishing regimes, as well as by patterns of primary productivity, and the delivery of land-based sources of nutrients and organic material. Some excellent, synoptic data sets exist for allof these aspects of the Scotian shelf environment (e.g. Elsner, 1999; O’Boyle, et al., 2000; andreferences therein).

It is important to reiterate that neither the deductive (primarily physical) nor inductive (primarilybiological) approaches require a theoretical understanding of the mechanistic controls over thedistribution and abundance of benthos. Correlation and ordination do NOT demonstratecausality, even at the level of the sediment-grain-size vs. infauna relationship (Snelgrove andButman, 1994), yet the empirical results of sound analyses can be used to accurately identifycommunities of organisms that are coherent in space and persistent in time.

In short, the deductive approach suffers from a lack of testing of the relationships betweenphysical forcing and community structure, while the inductive approach suffers from lack of data.It is for this reason that both approaches are generally combined in comprehensive benthicclassification schemes (Watson, 1997; Buchanan and Christian, 1999).

Recommendations for Developing a Classification of Benthic Ecosystem Types for theScotian Shelf

A set of fundamental conclusions and underlying assumptions are derived from the researchdescribed above, and presented as observations and recommendations to inform themethodology and outputs of the workshop exercise of selecting a classification scheme for thebenthic ecosystems of the Scotian Shelf. These are stated clearly and explicitly for debate.Elaborations and revisions resulting from the workshop will form the basis of further discussion,and may substantially alter the outcomes of the research.

1. The main purpose of the classification scheme is management of our oceans.

It is not necessarily the same as classification for scientific understanding. The application goal ofan ecosystem classification scheme is to guide the development an enforceable zoning schemeto control human activities on the shelf that have the potential to reduce or degrade marinebiodiversity.

Better understanding of how the Scotian Shelf ecosystem functions may be an important by-product of the classification exercise in the context of experimentation within adaptivemanagement regimes (sensu Holling, 1978). Hence, the classification scheme to be developedhere is a model for management and control, rather than for understanding and prediction(Bradbury, et al., 1983). The focus on precautionary and adaptive management has importantimplications for research priorities and budgets.

It is accepted that habitat classification schemes must be designed to meet a specific requirementor answer a well-defined question (Ray, 1976; Watson, 1997), for example definition ofbiodiversity hotspots. There can be no universal classification scheme for the Scotian Shelf thatmeets all possible requirements.

Page 25: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

24

2. SCALES of benthic classification should be initially set by compliance andenforcement capacity.

The operational limitations inherent in the implementation of zoning-based management on theScotian Shelf define the minimum spatial and temporal scales of resolution required forecosystem classification.

Zoning-based management in the present context depends on the resource users’ and themanagers’ ability to detect and respect spatially defined zones of activity in the open ocean.Given present technological and operational limitations, this resolves to a minimum manageableunit (MMU) of an area of approximately 100km2, with no more than 6 way points connected bystraight lines (J. Calvesbert, pers. comm.). It follows that the mapping of ecological units at finerresolution can be justified only if essential for the classification of benthic ecosystems at thespatial scales at which zoning can be implemented (e.g. Hatcher, et al., 1990). This practicalreality has important implications for the design and investment in sampling programmes aimedat mapping benthic biodiversity on the Scotian Shelf. The most common spatial scale of theminimum mapable unit used in benthic classification schemes developed for zoning is on theorder of 1 to 10 km2.

3. Classify ecosystems hierarchically and zone adaptively from the upper levels of thehierarchy.

The organization of nature is hierarchical, as is the organization of the societal structures thatlead to sound oceans governance. Both the spatial and temporal scales of seascapes, habitats,oceanographic and ecological processes form nested clusters rather than smooth continua(Dethier, 1990; Frith, et al., 1994). There simply does not exist a body of empirical knowledge or aset of theoretical models that can unequivocally identify the best classification scheme fordescribing and maintaining the diversity of benthic community types on the Scotian Shelf. Thegoal of optimally matching the tasks of zoning and monitoring to the levels and scales of naturalphenomena will only be achieved through a series of management experiments, the results ofwhich inform the subsequent trials. It is logical therefore to start this process at the higher levelsof the hierarchies (e.g. physiographic zones) where current knowledge is best, where cost-effectiveness is greatest, and where the risk of having to change the framework of classification islowest. Hierarchical classification schemes are well-suited to the addition of lower levels asknowledge increases, but are less amenable to the addition or removal of classes within levelsthan are flat classification systems (e.g. MarLIN; Hiscock, 1995), because of upwards-downwardslinkages and their implications for nesting.

4. Use the biological COMMUNITY as the minimum ecological unit (MEU) and the benthicHABITAT associated with the community as the minimum mapable unit (MMU) for thepurposes of management.

The biological community is the most appropriate lowest level of classification for characterizingbenthic biodiversity on the Scotian Shelf for three reasons: i) the high fidelity of species to benthiccommunities; ii) the close correspondence of benthic communities to habitats characterizable byattributes of the physical-chemical environment; and iii) the match between characteristic scalesof benthic habitats and management zones (i.e. MMUs).

Ecosystems may be conceptualized as comprising dual hierarchies of organization and functionthat do not map linearly onto spatial or temporal scales (O’Neill, 1989). Biodiversity is inherent atall levels in these hierarchies, not only at the level of the species population (Wilson, 1988). Thescience of marine ecology focuses primarily at the level of the community, and that is where ourmost robust paradigms, models and data sets apply (Bertness, et al., 2001). Biologicalcommunities map directly onto, and often define habitats, which are seen as the template ofecology (Southwood, 1977; Gray, 1997). Given the spatial scale of management and currentknowledge of species’ distributions on the Scotian Shelf, it is impractical to attempt mappingbiodiversity at the level of the species population. Even were that possible, the emergent

Page 26: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

25

properties that define enduring or recurrent features of community structure and function areunlikely to emerge from such an empirical classification (Watson, 1997; Buchanan and Christian,1999; Day and Roff, 2000; CEIMATE, 2000).

The main ecosystem services provided by the Scotian Shelf are the result of intact and functionalcommunities of organisms, rather than of their discrete populations. It follows that the first attemptto define indicators and set reference points for biodiversity preservation on the Scotian Shelfshould take a nested approach, starting with the integrated entity of the benthic community. Thecommunity focus has implications for the targets and methods of benthic sampling to beundertaken on the Scotian Shelf.

5. Incorporate the role of history explicitly into the classification of benthic communities.

The temporal dynamics of marine benthic communities dictate that the spatial patterns ofdistribution and abundance of organisms are not fixed, and future patterns cannot be predictedsolely on the basis of present patterns. The pattern of benthic communities currently observed onthe Scotian Shelf is not only a reflection of current environmental conditions, but also a temporalintegration of processes operating at multiple scales. At the scale of the shelf, biogeographicalprocesses related to ocean-climate systems, geological and evolutionary processes haveproduced cross- and along-shelf patterns of benthic communities. At the scale of benthichabitats, the cumulative flux of organic matter and the physical disturbance regime are theprimary determinants of the historical components of community pattern. As anthropogenicdisturbance is both extensive and patchily intense on the Scotian Shelf, a classification thatignores history may fail to identify benthic community types that were common or distinctive in theabsence of human impacts. Answers to the question whether we wish to maintain formerlycommon but currently benthic community types will have significant implications for the zoningmanagement regime to be implemented.

6. Equate habitat complexity to biodiversity in developing indicators and referencepoints.

Surrogate or proxy measures of species richness and community structure must be used to mapand monitor biodiversity at shelf scales because of the high costs of direct sampling and longterm monitoring. This is the accepted approach adopted by virtually every one who hasconsidered the problem (Cowardin, et al., 1979; Brown, 1993; Hiroven, et al., 1995; Zachariasand Howes, 1998; Day and Roff, 2000; and see reviews by Frith, et al., 1993; Watson, 1997;Buchanan and Christian, 1999; CEIMATE, 2000). The physical environmental indicators shouldbe synoptic and cost-effective for identifying the location, size and shape of areas that will bezoned to preserve the diversity of benthic community types. Remote sensing technologies are theonly way to accomplish this, and have been the method of choice for terrestrial habitat mappingsince the 1940s (CEIMATE, 2000). In the first order, indicators of habitat complexity may be usedto predict areas of high biodiversity and to maximize the range of benthic community typesincluded within a management zones. Verification and refinement by direct sampling of biotashould be driven by hypotheses about the mechanistic nature of relationships between physicalcontrols and biological processes generated from current theory and from prospective surveysand analyses.

Empirical correlations have been established in many ecosystems between physical-chemicalattributes of marine habitats and the structure and function of communities of organisms thatinhabit them (e.g. Warwick, 1988; McGhee, 1994; Hiscock, 1995; Mahon, et al., 1998; Kostylev,et al., 2001). Such relationships are rarely robust at the level of the individual species’distribution, primarily because of temporal variability in disturbance regimes, recruitment andmovement (Dayton, 1971; Underwood and Fairweather, 1989; Sousa, 2001). At the scale ofzoning-based management (i.e. 10s of kilometres), habitat complexity includes the variety ofhabitats within an area as well as the patchiness of those habitats (i.e. the mosaic grain, sensuLevin, 1992), and the distances between them (i.e. connectedness). Within habitat patches, therugosity (i.e. fine-scale topographic complexity, including sediment grain size) is the most

Page 27: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

26

powerful indicator of community structure. While biota create habitat in many marineenvironments (e.g. Woodin, 1978; Thistle and Eckman, 1990), such “foundation species” aregenerally confined to shallow, nearshore, euphotic areas (Bruno and Bertness, 2001). In somehabitats on the Scotian Shelf, certain bivalves and corals may serve as foundation species.Among habitat patches (i.e. at the larger scales of the Scotian Shelf), physical-chemicalenvironmental variates related to the geological (e.g. bedforms) and oceanographic (e.g. thermaland hydrodynamic) regimes exert the most predictable control on the diversity of benthiccommunity types.

It is one thing to identify relationships between physical attributes of an environment and thebiotic components (e.g. the increase in filter feeders with near-bed flow velocity, Snelgrove andButman, 1994). It is quite another to use the relationships to predict the space-time co-ordinatesof maximum abundance and diversity of organisms from those relationships. The interplaybetween ecosystem structure (in terms of biodiversity) and function is far from obvious or linear interms of the relative effects of physical and biological controls (Mooney, et al., 1996). Thisambiguity begs the question of whether causal or mechanistic relationships between physicalforcing and biological process are necessary and sufficient for predicting benthic communitystructure.

Conclusions

1. The DFO should take a hierarchical, risk-averse approach to the classification process. (i.e.the methodology must mirror the structure and organization of nature). Start at high levels ofecological organization using broad scale mapping of variates selected on robust theory andempirical relationships, then working down to lower levels and smaller domains as required.

2. It is not presently possible, and will likely never be possible to delineate the exact boundariesof marine communities on the Scotian shelf. It is possible, however, for the purpose ofclassification, to delineate the boundaries of areas approximating or exceeding minimummanagement units (approx. 100km2) that have a very high probability of containingrepresentative or unique assemblages of benthic organisms.

3. A minimum set of physical-chemical parameters can be specified a priori that already exist inthe literature or available data banks, that will provide defensible biodiversity criteria fordemarcating MMUs, and practical guidelines for monitoring. The parameters that encompassthe known requirements of intact benthic communities and are known to influence biodiversityare: topographic and substratum complexity, intermediate disturbance regimes, predictabledelivery of nutrients, organic material and reproductive propagules. A set of 6 to 10measurement variates are adequate for a first set on indicators and reference points formaintaining the diversity of ecosystem types on the Scotian shelf.

References

Andrefouet, S., M. Claereboudt, P. Matsakis, J. Pages, and P. Dufour. 2001. Typology of atollrims in Tuamotu archipelago (French Polynesia) at landscape scale using SPOT HRVimages. Int. J. Remote Sensing 22:987-1004.

Andrew, N.L., and A.L. O’Neill. 2000. Large-scale patterns in habitat structure on subtidal rockyreefs in New South Wales. Mar. Freshw. Res. 51:255-263.

Andrew, N.L., and B.D. Mapstone. 1987. Sampling and the description of spatial pattern inecology. Ann. Rev. Oceanogr. Mar. Biol. 25:39-90.

Auster, P.J. 1998. A conceptual model of the impacts of fishing gear on the integrity of fishhabitats. Conserv. Biol. 12:1198-1203.

Page 28: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

27

Auster, P.J., C, Michalopoulos, P.C. Valentine, and R.J. Malatesta. 1998. Delineating andmonitoring habitat management units in a temperate deep-water marine protected area;pp. 169-185. In: N.W. Munro and J.H.M. Willison (eds.), Linking protected areas withworking landscapes, conserving biodiversity. Science and Management of ProtectedAreas Association, Wolfville, Nova Scotia.

Beier, P. 1993. determining minimum habitat areas and habitat corridors for cougars. ConservBiol. 7:94-108.

Belsky, M.H. 1989. Developing an ecosystem management regime for large marine ecosystems;pp. 443-460. In: K. Sherman and L.M. Alexander (eds.), Biomass yields of large marineecosystems. Amer. Assoc. Adv. Sci. Symp. 111, Washington, DC.

Bertness, M.D., S.D. Gaines, and M.E. Hay. 2001 Marine community ecology. Sinauer Assoc.,Mass.; 550p.

Booth, J., D.E. Hay, and J. Truscott. 1996. Standard methods for sampling resources andhabitats in coastal subtidal regions of British Columbia: Part 1: review of mapping withpreliminary recommendations. Can. Tech. Rep. Fish. Aquat. Sci. 2118; 53p.

Bradbury, R.H., R.E. Reichelt, and D.G. Green. 1983. Explanation prediction and control in coralreef ecosystems. III. Models for control. Proc. Great Barrier Reef Conf. Audst. Inst. Mar.Sci., Townsville, Qld:165-170.

Brown, B. 1993. A classification system of marine and estuarine habitats in Maine: An ecosystemapproach to habitats Part I: benthic habitats. Marine Natural Areas Program; 51p +Append.

Brown, D. 1998. Habitat essential for sustainable fisheries. J. Shellfish Res. 17:1481-1482.

Brown, S.K., K.R. Buja, S.H. Jury, and M.E. Momnaco. 2000. Habitat suitability index models foreight fish and invertebrate species in Casco and Sheepscot bays, Maine. NA. J. Fish.Manag. 20:408-435.

Bruno, J.F., and M.D. Bertness. 2001. Habitat modification and facilitation in benthic marinecommunities; pp. 201-218. In: M.D. Bertness, S.D. Gaines, and M.E. Hay (eds.), MarineCommunity Ecology. Sinauer Assoc. Sunderland, Mass.

Buchanan, R.A., and J. Christian. 1999. Marine habitat classification – Literature review. DFOMarine Environment and Habitat Management Division, St. John’s, Nfld.; 37p. + 4Append.

Butler, M.J.A., C. LeBlanc, J.A. Belbin, and J.L. MacNeill. 1987. Marine resource mapping: anintroductory manual. FAO Fish. Tech. Pap. No. 274; 256p.

Caddell, S.E. 1998. Application of an acoustic sea floor classification system for benthic habitatassessment. J. Shellfish Res. 17:1459-1461.

Canessa, R.R. 1998. GIS decision support for integrated coastal management. Oceans Conserv.Rep. Ser., DFO, Axys Environ. Consult. Sidney, BC; 176p.

Catto, N.R., R. Hooper, M.R. Anderson, D.A. Scruton, L.M.N. Ollerhead, and U.P. Williams. 1997.Biological and geomorphological classification of Placentia Bay: a preliminaryassessment. Can. Tech. Rep. Fish. Aquat. Sci. No. 2842; 49p.

Page 29: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

28

CEIMATE. 2000. Ecological indicators for the nation. Committee to Evaluate Indicators forMonitoring Aquatic and Terrestrial Environments, National Academy Press, WashingtonDC; 148p + 3 Append.

Clarke, K.R., and M. Ainsworth. 1993. A method of linking multivariate community structure toenvironmental variables. Mar. Ecol. Prog. Ser. 92:205-219.

Connor, D.W., M.J. Dalkin, T.O. Hill, R.H.F. Holt, and W.G. Sanderson. 1997. Marine biotopeclassification for Britain and Ireland. Vol. 1. Littoral biotopes. Joint Nature ConservationCommittee, Peterborough, JNCC Report No. 229, Version 97.06. Joint NatureConservation Committee, Peterborough, JNCC Report No. 230, Version 97.06.

Connor, D.W., M.J. Dalkin, T.O. Hill, R.H.F. Holt, and W.G. Sanderson. 1997. Marine biotopeclassification for Britain and Ireland. Vol. 2. Sublittoral biotopes. Joint NatureConservation Committee, Peterborough, JNCC Report No. 230, Version 97.06. JointNature Conservation Committee, Peterborough, JNCC Report No. 230, Version 97.06..

Costello, M.J. 2000. Developing species information systems: the European registrar of marinespecies (ERMS). Oceanography 13:48-55.

Cowardin, L.M., V. Carter, F.C. Golet, and E.T. LaRoe. 1979. Classification of wetlands anddeepwater habitats of the United States. U.S. Dept. of the Interior, Fish and WildlifeService, FWS/OBS-79/31; 103p.

Cowen, R., R.K. Cowen, K.M.M. Lwiza, S. Spnaugle, C.B. Paris, and D.B. Olsen. 2000.Connectivity of marine populations: Open or closed? Science 287:857-859.

Coyne, M.S., M.E. Monaco, M. Anderson, W. Smith, and P. Jokiel. 2001. Hawaii benthic habitatclassification scheme for mapping of main eight Hawaiian islands. BiogeographyProgram. U.S. Dept. Comm. NOAA National Ocean. Service, Silver Springs, MD; 14p.

Cullinan, V.I., and J.M. Thomas. 1992. A comparison of quantitative methods for examininglandscape pattern and scale. Landscape Ecol. 7:211-227.

Davies, C.E., and D. Moss. 1999. EUNIS Habitat Classification. Final Report to the EuropeanTopic Centre on Nature Conservation. European Environment Agency, Paris; 256p.

Davis, D.S., and S. Browne (eds.). 1994. Natural history of Nova Scotia (2nd Edition). Province ofNova Scotia, Halifax, NS; 498p.

Davis, D.S., P.L. Stewart, R.H. Loucks, and S. Browne. 1994. Development of a biophysicalclassification of offshore regions for the Nova Scotia continental shelf. Proc. CoastalZone Canada Conf., Halifax, NS:2149-2157.

Day, J., and J.C. Roff. 2000. Planning for representative marine protected areas: A framework forCanada’s oceans. World Wildlife Fund Canada, Toronto; 147p.

Dayton, P.K. 1971. Competition, disturbance and community organization: the provision andsubsequent utilization of space in a rocky intertidal community. Ecol. Monogr. 41:351-389.

Deither, M.N. 1990. A marine and estuarine classification system for Washington State. Dept.Natural Resources; 56p.

Dempsey, J., A. Simms, J. Harper, E. Lambert, and R. Hooper. 1995. West coast Newfoundlandoil spill sensitivity atlas Environ. Stud. Res. Fund Rep. No. 127,Calgary; 62p.

Page 30: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

29

Elsner, P. 1999. The Scotian shelf Seascape: Using GIS for quantitative resource evaluation andmanagement – a case study. M.M.M. Thesis, Faculty of Graduate Studies, DalhousieUniversity, Halifax, Nova Scotia; 94p.

ECC. 2000. Marine, coastal and estuarine investigations: Final report. Environ Conserv Council,Victoria; 50p. + 6 Append. URL: www.nre.vic.gov.au.

ECL. 1999. Review of terms and definitions relevant to ocean and marine resource managementunder The Ocean Act. Environmental Consultants Ltd. Report to Dept. Fisheries andOceans Canada, Ottawa; 9p.

ESC. 2000. Millennium ecosystem assessment: strengthening the capacity to managementecosystems for human development – Project overview. Exploratory Steering Committee,Seattle; 20p. + 3 Append.

Fader, G.B.J., R.A. Pickrill, B.J. Todd, R.C. Courtney, and D.R. Parott. 2000. The emerging roleof marine geology in benthic ecology. Geol. Survey Can. Tech. Pap. 00-16:1-9.

Farina, A. 1998. Principles and methods in landscape ecology. Chapman and Hall, London; 89p.

Frank, K.T., and W.C. Leggett. 1994. Fisheries ecology in the context of ecological andevolutionary theory. Annu. Rev. Ecol. Syst. 25:401-422.

Frith, H.R., G. Searing, and P. Wainwright. 1994. Methodology for a B.C. shoreline biotic mappingsystem. LGL Ltd. Rep. to B.C. Environment, Lands and Parks; 38p.

Frith, H.R., G. Searing, P. Wainwright, J. Harper, and B. Emmett. 1993. Review of habitatclassification systems and an assessment of their suitability to coastal B.C. LGL Ltd.Rept. to B.C. Environment, Lands and Parks; 87p.

Gomes, G. 1998. Review of the NOAA National coastal management programs for estuary andcoastal wetland protection. IEC, Oakville, ON; 66p.

Gomes, M.C., R.L. Haedrich, and J.C. Rice. 1992. Biogeography of groundfish assemblages onthe Grand Bank. J. Northwest Atlantic Fish. Sci. 14:13-27.

Greene, H.G., M.M. Yoklavich, R.M. Starr, V.M. O’Connell, W.W. Wakefield, D.E. Sullivan,J.E. McRae, and G.M. Cailliet. 1999. A classification scheme for deep seafloor habitats.Oceanologica Acta 22:663-678.

Gregory, D., B. Petrie, F. Jordan, and P. Langille. 1993. Oceanographic, geographic, andhydrological parameters of Scotia-Fundy and southern Gulf of St. Lawrence inlets. Can.Tech. Rep. Hydrogr. Ocean. Sci. 142:248p.

Gray, J.S. 1997. Marine biodiversity: patterns, threats and conservation needs. Biodiversity andConservation. 6:153-175.

Harper, J.R., G.A. Robilliard, and J. Lathrop. 1983. Marine regions of Canada: framework forCanada’s System of National Marine Parks. Parks Canada, Ottawa; 49p.

Harper, J.R., J. Christian, W.E. Cross, R. Frith, G. Searing, and D. Thompson. 1993.Classification of the marine regions of Canada. Environment Canada, Ottawa; 77p.

Harrison, W.G., and D.G. Fenton (eds.). 1998. The Gully: A scientific review of its environmentand ecosystem. Can. Stock Asses. Sect. Res. Doc. 98/93; 282p.

Page 31: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

30

Hatcher, A.I., G.D. Wright, and B.G. Hatcher. 1990. Resolving the conflict between conservationvalues and extractive use of the Abrolhos coral reefs. Proceedings of the EcologicalSociety of Australia 16:55-70.

Hatcher, B.G. 1997. Coral reef ecosystems: how much greater is the whole than the sum of theparts? Coral Reefs 16(Suppl.):77-91.

Hatcher, B.G., J. Imberger, and S.V. Smith. 1987. Scaling analysis of coral reef systems: Anapproach to problems of scale. Coral Reefs 5:171-181.

Hedgepeth, J.W. (ed.). 1957. Marine biogeography; pp. 359-382. In: J.W. Hedgepeth (ed.),Treatise on marine ecology and paleoecology. Mem. Geol. Soc. Amer. 67; 285p.

Helsinki Commission.1998. Red list of marine and coastal biotopes and biotope complexes of theBaltic Sea, Belt Sea and Kattegat - including a comprehensive description andclassification system for all Baltic marine and coastal biotopes. Baltic Sea EnvironmentProceedings No. 75, Helsinki; 136p.

Hiroven, H.E., L. Harding, and J. Landucci. 1995. A national marine ecological framework forecosystem monitoring and state of the environment reporting; pp.117-129. In:N.L. Skackell and J.H.M. Willison (eds.), Marine protected areas and sustainablefisheries. Acadia Univ. Press, Wolfville, NS.

Hiscock, K. (ed.). 1995. Classification of benthic marine biotopes of the North-east Atlantic. Proc.BioMar Workshop, Cambridge. Peterborough Joint Nature Conservation Committee;105p.

Hiscock, K. 1999. The Marine Life Information Network for Britain and Ireland. MarLIN News 1:1-7. URL: www.marlin.uk.

Holling, C.S. 1978. Adaptive environmental assessment and management. Wiley InterScience,Chichester.

Holthus, P.F., and J.E. Maragos. 1995. Marine ecosystem classification for the tropical islandPacific; pp 239-278. In: J.E. Maragos, M.N.A. Peterson, L.G. Eldredge, J.E. Bardach, andH.F. Takeuchi (eds.), Marine and coastal biodiversity in the tropical island Pacific region.East-West Center, Honolulu.

Hooper, R.G. 1997. A general coastal marine habitat classification system for Newfoundland.Rept. to Fisheries and Oceans, St. John’s, Nfld. 14p.

Hyrenbach, K.D., K.A. Forney, and P.K. Dayton. 2000. Marine protected areas and ocean basinmanagement. Aquat. Conserv.: Mar. Freshw. Ecosyst. 10:437-458.

ICES. 2000. Report of the study group on ecosystem assessment and monitoring. ICES CM2000/E:09, Marine Habitat Committee, Intl. Counc. Explor. Sea, Copenhagen; 38p.

Jax, K., C.G. Jones, and T.A. Pickett. 1998. The self-identity of ecological units. Oikos 82:253-264.

Kostylev, V.E., B.J. Todd, G.B.J. Fader, R.C. Courtney, G.D.M. Cameron, and P.A. Pickrill. 2001.Benthic habitat mapping on the Scotian Shelf based on multi-beam bathymetry, surficialgeology and sea floor photographs. Mar. Ecol. Prog. Ser. 219:121-137.

Legendre, P., and L. Legendre. 1998. Numerical Ecology. Elsevier, Amsterdam; 281p.

Levin, S.A. 1992. The problem of pattern and scale in ecology. Ecology 73:1943-1967.

Page 32: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

31

Levinton, J.S. 1982. Marine Ecology. Prentice-Hall, NJ; 392p.

Loder, J.W., J.A. Shore, C.G. Hannah, and B.D. Petrie. 2001. Decadal-scale hydrographic andcirculation variability in the Scotia-Maine region. Deep-Sea Res. 48:3-35.

Longhurst, A.R. 2001. The ecological geography of the sea. Academic Press, New York.

Loehle, C., and G. Wein. 1994. Landscape habitat diversity: A multi-scale information theoryapproach. Ecol Modelling 73:311-329.

Lyne, V., and P. Last. 2001. Interim marine bioregionalization for Australia – Towards a nationalsystem of marine protected areas. CSIRO, Hobart, 87p. URL:[email protected].

Mahon, R., S.K. Brown, K.C.T. Zwanenburg, D.B. Atkinson, K.R. Buja, L. Claflin, G.D. Howell,M.E. Monaco, R.N. O’Boyle, and M. Sinclair. 1998. Assemblages and biogeography ofdemersal fishes of the east coast of North America. Can. J. Fish. Aquat. Sci. 55:1704-1738.

Mann, K.H., and J.R.N. Lazier. 1996. Dynamics of marine ecosystems: biological-physicalinteractions in the Oceans. Blackwell Science, London; 257p.

Massell, S.R., and T.J. Done. 1993. Effects of cyclone waves on massive coral assemblages onthe Great Barrier reefmeterology, hydrodynamics and demography. Coral Reefs 12:153-166.

McGhee, M.A. 1994. Correspondence between assemblages of coral reef fishes and gradients inwater motion, depth and substrate size off Puerto Rico. Mar. Ecol. Prog. Ser. 105:243-255.

MEQAG. 1994. Marine Ecological Classification System for Canada. Marine EnvironmentalQuality Advisory Committee, Environment Canada, Ottawa; 21p.

Miller, R.I. (ed.). 1994. Mapping the diversity of nature. Chapman and Hall, Melbourne; 341p.

Minns, C.K. 1997. Science for fish habitat management: a fresh perspective. Can. Tech. Rept.Fish. Aquat. Sci. 2147:81-93.

Mooney, H.A., J.H. Cushman, E. Medina, O.E. Sala, and E.-D. Schulze (eds.). 1996. Functionalroles of Biodiversity: A global Perspective. Scientific Committee on Problems in theEnvironment (SCOPE), J. Wiley and Sons, New York:393-429.

Muldoon, J. (ed.). 1995. Towards a marine regionalization for Australia. CSIRO Workshop Rept.March 1994; 243p.

Mumby, P.J. 2001. Habitat and beta diversity of marine systems: a new approach tomeasurement scaling and interpretation. Oecologia (in press).

Mumby, P.J., and A.R. Harborne. 1999. Development of a systematic classification scheme ofmarine habitats to facilitate management and mapping of Caribbean coral reefs. Biol.Conserv. 88:155-163.

NAS. 2000. Ecological indicators for the nation. National Academic Press, Washington, DC.

Page 33: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

32

NCER. 2000. Development of national aquatic ecosystem classifications and referenceconditions. Natl. Center Environ. Res., U.S. Environmental Protection Agency,Washington, DC; 7p. URL: http://es.epa.gov/ncerqa.html.

O’Boyle. R. (ed.). 2000. Proceedings of a workshop on the ecosystem considerations for theEastern Scotian Shelf Integrated Management (ESSIM) Area. CSAS Proc. Series2000/14; 92p.

O’Neill, R.V. 1989. Perspectives in hierarchy and scale; pp. 140-156. In: J. Roughgarden,R.M. May, and S.A. Levin (eds.), Perspectives in ecological theory. Princeton Univ.Press, New Jersey.

O’Neill, R.V., D.L. DeAngelis, J.B. Waide, and T.F.H. Allen. 1986. A hierarchical concept ofecosystems Princeton Univ. Press, NJ; 253p.

Page, F.H., and R.J. Losier. 1994. Temperature variability during Canadian bottom-trawl summersurveys conducted in NAFO division 4VWX, 1970-1992. ICES Mar. Sci. Symp. 198:323-331.

Pascual, M., and S.A. Levin. 2000. Spatial scaling in a benthic population model with density-dependent disturbance. Theoret. Pop. Biol. 56:106-122.

Percy, J.A., P.G. Wells, and A.J. Evans (eds.). 1996. Bay of Fundy issues: a scientific overview.Environ .Can. Atl. Reg. Occ. Rept. No. 8; 191p.

Petrie, B., P. Yeats, and P. Strain. 1999. Nitrate, silicate and phosphate atlas for the Scotian shelfand the Gulf of Maine. Can. Tech. Rept. Hydrogr. Ocean. Sci. 203; 96p.

Pielou, E.C. 1972. Measurement of structure in animal communities; pp. 341-352. In: J.A. Wiens(ed.), Ecosystem structure and function. Oregon Univ. Press, Corvallis.

Platnick, N.I. 1991. Patterns of biodiversity: tropical vs. temperate. J. Natur. Hist. 25:1083-1088.

Polis, G.A., M.E. Power, and G.R. Huxel (eds.). 2001. Food webs at the landscape level. Univ.Chicago Press, Ill (in press).

Rapport, D.J., C.L. Gaudet, and P. Calow (eds.). 1995. Evaluating and monitoring the health oflarge-scale ecosystems. Springer, Berlin; 457p.

Rapport, D.J., R. Costanza, and A.J. McMichael. 1998. Assessing ecosystem health. TREE13:397-402.

Ray, G.C. 1976. Critical marine habitats: definition, description, criteria and guidelines foridentification and management. IUCN, Morges, Switzerland; 50p.

Ricklefs, R.E., and D. Schluter (eds.). 1993. Species diversity in ecological communities. Univ.Chicago Press, Chicago; 414p.

Robinson, C.L.K., and C.D. Levings. 1995. An overview of habitat classification systems,ecological models, and geographic information systems applied to shallow foreshore andmarine habitats. Can. Manusc. Rep. Fish. Aquat. Sci. 2322; 65p.

Robinson, C.L., D.E. Hay, J. Booth, and J. Truscott. 1996. Standard methods for samplingresources and habitats in coastal subtidal regions of British Columbia: Part 2: review ofsampling with preliminary recommendations. Can. Tech. Rep. Fish. Aquat. Sci. 2119;119p.

Page 34: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

33

Roff, J.C., and M.E. Taylor. 2000. National frameworks for marine conservation – a hierarchicalgeophysical approach. Aquat. Conserv. Mar. Freshw. Ecosyst. 10:209-223.

Sinclair, M., R. O’Boyle, L. Burke, S. D’Entremont. 1999. Incorporating ecosystem objectiveswithin fisheries management plans in the Maritimes region of Atlantic Canada. ICES CM1999/Z03; 20p.

Sheppard, C.R., K. Matheson, J.C. Bythel, P. Murphy, C.B. Myers, and B. Blake. 1998. Habitatmapping in the Caribbean for management and conservation: Use and assessment ofaerial photography. Aquat. Cons. 5:277-298.

Singh, R., M.-I. Buzeta, M. Dowd, J.L. Martin, and M. LeGresley. 2000. Ecological overview ofMusquash estuary: A proposed marine protected area. Can. Manu. Rept. Fish. Aquat.Sci. No. 2538; 39p.

Snelgrove, P.V.R. 1999. Getting to the bottom of marine biodiversity: sedimentary habitats.BioScience 49:129-138.

Snelgrove, P.V.R., and C.A. Butman. 1994. Animal-sediment relationships revisited: causeversus effect. Oceanogr. Mar. Biol. Ann. Rev. 32:111-177.

Sousa, W.P. 2001. Natural disturbance and the dynamics of marine benthic communities;pp. 201-218. In: M.D. Bertness, S.D. Gaines, and M.E. Hay (eds.), Marine CommunityEcology. Sinauer Assoc. Sunderland, Mass.

Steele, J. 1989. The ocean “Landscape”. Landscape Ecol. 3:185-192.

Stewart, P.L., H.A. Levy, and B.T. Hargrave. 2001. Database of benthic biomass and productivitymeasurements for the eastern continental shelf, slope and adjacent areas. Can. Tech.Rept. Fish. Aquat. Sci. (in press); 27p. + Append.

Strong, D.R., D. Simberloff, L.G. Abele, and A.B. Thistle (eds.). 1984. Ecological communities:conceptual issues and the evidence. Princeton Univ. Press, Princeton, NJ.

Sullivan Sealy, K., and G. Bustamante. 1999. Setting priorities for marine conservation in LatinAmerica and the Carbbean. The nature Conservancy, Arlington, VA; 125p.

Thistle, D., and J.E. Eckman. 1990. The effect of a biologically produced structure on the benthiccopepods of a deep-sea site. Deep Sea Res. 37:541-554.

Underwood, A.J., and P.G. Fairweather. 1989. Supply-side ecology and benthic marineassemblages. Trends Ecol. Evol. 4:16-20.

UNEP. 1998. Annex III. Revised classification of benthic marine habitat types for theMediterranean region. In: WG. 149/5: Report: Meeting of experts on marine habitat typesin the Mediterranean region. SPA/RAC, Tunis.

Warwick, R.M. 1988. Analysis of community attributes of the macrobenthos ofFrierfjord/Langesundfjord at taxonomic levels higher than species. Mar. Ecol. Prog. Ser.46:167-170.

Warwick, R.M., and R.J. Uncles. 1980. Distribution of benthic macrofauna associations in theBristol Channel in relation to tidal stress. Mar. Ecol. Prog. Ser. 3:97-103.

Warwick, R.M., and K.R. Clarke. 1998. Taxonomic distinctness and environmental assessment.J. Appl. Ecol. 35:532-543.

Page 35: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

34

Watson, J. 1997. A review of ecosystem classification: delineating the Strait of Georgia. DFOScience Branch, Pacific Region, Vancouver, BC; 81p.

Wiens, J.A. 1989. Spatial scaling in ecology. Func. Ecol. 3:385-397.

Wilken, E.B., D. Gauthier, I. Marshall, K. Lawton, and H. Hiroven. 1996. A perspective onCanada’s ecosystems. An overview of terrestrial and marine ecozones. CanadianCouncil of Ecological Areas Occasional Paper No. 14.

Wilson, E.O. (ed.).1988. Biodiversity. National Academy of Sciences/Smithsonian Institution,Washington, DC; 538 p.

Williams, G.L. 1989. Coastal/estuarine fish habitat description and assessment manual – Part II.Habitat description procedures. Dept. Fisheries and Oceans, Vancouver, BC; 38p. +Append.

Wolanski, E. (ed.). 2000. Oceanographic processes of coral reefs: physical and biological links inthe Great Barrier Reef. CRC Press, Baton Rouge.

Woodin, S.A. 1978. Refuges, disturbance and community structure: a marine soft-bottomexample. Ecology 59:274-284.

Zacharias, M.A., and D.E. Howes. 1998. An analysis of marine protected areas in BritishColumbia, Canada, using a marine ecological classification system. Nature Areas Journal18:4-9.

Zacharias, M.A., D.E. Howes, J.R. Harper, and P. Wainwright. 1998. The British Columbia marineecosystem classification: rationale, development, and verification. Coastal Management26:105-124.

Zwanenburg, K.C.T. 2000. The effects of fishing on demersal fish communities of the ScotianShelf. ICES J. Mar. Sci. 57:503-509.

Page 36: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

35

SUMMARY OFWORKSHOP PRESENTATIONS

Page 37: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

36

AN OCEANS MANAGEMENT PERSPECTIVE ONA BENTHIC CLASSIFICATION FRAMEWORK

Robert RutherfordOceans and Coastal Management DivisionDFO Maritimes RegionDartmouth, Nova Scotia

The purpose of this presentation is to highlight the need and role for an ecological, includingbenthic, characterization framework for oceans management and planning in the Scotia-Fundyarea. As one of the key “end users” of current frameworks, any future product from this exercise,we have provided a brief overview of past approaches used for characterization in the offshoreportion of the Scotia-Fundy region; examined the application and key elements of a future benthicframework; and provided recommendations on next steps.

The Oceans Act: Basis for Ecosystem Approach to Ocean Management and Planning

The Oceans Act provides a legislative foundation for integrated oceans and coastal managementin Canada’s estuarine, coastal and marine ecosystems. For the purposes of this workshop, it isuseful to note that the Act mandates DFO, in collaboration with other organizations and groups, toadopt and operationalize the sustainable development, ecosystem and precautionary approachesfor the management and conservation of our marine environments. This means balancingmultiple oceans use and economic development with the conservation and protection ofbiodiversity and marine health at all ecosystem levels. The Act also recognizes that effectivemanagement requires a foundation of scientific understanding of ecosystems and all theircomponents. Therefore, the development of both a broad ecosystem and a benthiccharacterization and conservation framework is essential for the implementation of the OceansAct.

Fisheries and Oceans Canada is implementing the vision of the Oceans Act through severalprograms for the management and protection of coastal and offshore environments in theMaritimes. Individual initiatives conducted over the past few years are being implemented onwide range of spatial scales. At the broadest scale, the region has been divided into two LargeOcean Management Areas (LOMAs), with the Eastern Scotian Shelf Integrated Management(ESSIM) initiative representing the primary effort to date. Within the LOMAs are a number ofCoastal Management Areas (CMAs). Projects include planning efforts in the Minas Basin in theBay of Fundy and the Bras d'Or Lakes. Two marine protected area (MPA) initiatives include theSable Gully and the Musquash Estuary. These are the first of a broader system of MPAs plannedfor the region.

Frameworks to Assist Planning

As part of many of these initiatives we are compiling “ecosystem overviews” to assist in theplanning activities, including descriptions of benthic communities. To complement theseinformation gathering exercises ecosystem objectives are being developed at both thegeneral/national scale and at the smaller scale through vision setting discussions with the localcommunities. Species habitat profiles for indicator or key species will be used with measurablehealth indicators (e.g. salinity, temperature ranges, and acceptable contaminant levels) to helpdefine the ecosystem vision. This ecosystem framework is part of the management plan’sperformance measurement, which tells us if we are moving toward the definedsocial/economic/environmental management vision. An agreed to benthic characterizationframework will assist in the identification of appropriate benthic indicators and managementtargets for the maintenance of the health, productivity and diversity of the ecosystem types.

From an ocean management and planning perspective, a benthic framework will provide afundamental ecological input to plans and decisions on how our ocean areas will be managedand used. Future ocean use zoning and MPA planning provide good examples of future

Page 38: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

37

applications for a benthic characterization framework, e.g. capturing particularly special andimportant benthic communities. Given the complexity of oceans management and decision-making, the benthic component is just one of many layers to consider when exploring zoningconcepts and options for an area such as the Scotian Shelf. It is just one, albeit an essential,piece of the puzzle. A useful example to consider from elsewhere is the work recently completedat Great Barrier Reef to assist with both long-term and day-to-day planning needs. It is notbenthic-specific and it provides a synthesis of many data layers. The Barrier Reef approachillustrates the applications of ecosystem classification frameworks to oceans management andplanning, essentially providing a basis for decision-making and ocean use zoning.

One of the key uses of the framework will be to provide direction for the identification ofecologically sensitive areas, and for the assessment of impacts/effects of individual and multipleactivities. As well, a benthic framework will be an important tool for the environmental assessmentprocess, although it must be emphasized that a regional framework will not remove the need forsite-specific assessments. It is hoped that the framework will also contribute to the scientificcommunity, providing a basis for ecosystem understanding, scientific inquiry and researchplanning.

What is Currently Available to Oceans Planners?

To support initiatives and broad objectives noted above what is currently available “off the shelf”to ocean planners and the various sectors and communities involved? Over the years severalapproaches by government and non-government researchers to describing benthic communitiesin the offshore. The various efforts include taking biological, biophysical, and physicalapproaches. The presentation illustrated samples of the various data collections, mappingapproaches and planning “tools” currently available to oceans planners to incorporate benthicconsiderations in oceans management.

A common biological approach is to use resource survey information for particular species, suchas scallops. Although this helps in the identification of key or important benthic organisms insome areas, non-commercial species are rarely captured in the analysis and the spatial coveragecorresponds to primary fishing grounds, e.g. bank environments. However, the fact that we havea good knowledge of commercial species will help us to both develop and test the benthicclassification scheme in the Scotia-Fundy area. For many stakeholders the planning objectivesset around commercial species will be highest importance. Unfortunately the level of samplingacross the shelf has been limited, and therefore restricts our ability to solely rely upon biologicaldata to develop the scheme.

Ecological “classification” has a long history in Canada - particularly at a national scale. Forexample, Parks Canada has developed the concept of marine regions, which was developed forNational Marine Conservation Area planning. This scale is not very useful for planning initiativessuch as ESSIM. On a Scotian Shelf scale the Nova Scotia Museum has taken a Natural historyapproach and derived a scheme showing basic biophysical zonation of the marine environment.Characteristic benthic organisms are provided for each zone. Although the spatial coverage ofthe scheme is adequate for ocean planning, the level of detail on the specific ecological orbenthic organisms is lacking. Similarly, in support of environmental assessments for the oil andgas industry a number of mapping efforts have been used to classify the benthos. Often thisapproach characterizes “dominant organisms” on the Scotian Shelf with the intent of viewing theimpacts and setting of specific projects.

In 1998, the WWF (Day and Roff) began work on developing a hierarchical model for marineclassification using the Scotian Shelf as a case study area. This approach combines variousphysical data layers to characterize the benthic habitat or ecotypes, as well as pelagiccomponents. It was felt that we had a better understanding of the physical variables that combineto make up the different ecotypes than we did of the biological or distribution of speciesassemblages. The theory is very appealing and should be seriously considered in anycharacterization scheme that is developed but it did not provide a map, which explains what we

Page 39: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

38

know of the biological distribution of assemblages. There are several reasons for this; DFO didnot have the right data to feed the model for example the surfical geology did not include theimmediate subsurface or the bed forms, which are both controlling factors; some variables whichwere applied on a large scale, for example temperature, had a controlling or limiting effect oncommunities at a smaller scale and range, so was under weighted in the characterization; tidalmixing along thermoclines on the Shelf edge produced well fed benthic communities that thestratification layer, used as a surrogate for mixing, implied did not exist; depth was not acontrolling factor as was assumed; and bottom currents which have a major affect on the benthoswere not available and so not included.

Although this map has several features marine biologists might recognize, it has raised manyquestions and does not reflect what we perceive to be the characterization based on ourunderstanding of the biology.

From the oceans planning perspective and to meet the objectives described earlier, there is noone “off the shelf” product that meets our needs, either from a spatial extent or a biologicaldescription perspective. This underscores the requirement for a synthesis of existing informationinto one agreed framework for classifying, understanding and conserving/managing benthiccommunities.

Conclusion

In conclusion, it is useful to put out some suggested “goals” or “criteria” for building the benthicframework:

• Beyond the requirement for solid, peer-reviewed science, broad agreement among thescience, management/regulatory, and marine user communities is essential for acceptance offuture decisions based on the framework.

• The framework will need to be geo-referenced so that it can be incorporated into variousmapping initiatives, a fundamental tool for ocean planning and decision-making.

• When building the framework, we have to consider the scales of use and application. Ahierarchical approach will allow for application through a range of spatial scales.

• Consider both infauna and epibenthic organisms in the scheme

• Bearing in mind that we will never be satisfied with the amount and quality of our informationand understanding, we have to adopt a pragmatic approach based on what we do know andwhat data we have, this will mean the use of predictive/deductive approaches based onexpert opinion and theory derived from the areas were we have enough data to take a datadriven approach. By building durability and flexibility into the framework, we can add, changeand improve it over time.

In recent years there has been growing agreement on the layers and variables (e.g. substrate,water column characteristics, temperature, currents, temporal circulation, depth, structuralcomplexity, level of natural disturbance, etc.). The challenge now lies with the details in selectingand combining these layers.

References

Day, J., and J.C. Roff. 2000. Planning for representative marine protected areas: A framework forCanada’s oceans. World Wildlife Fund Canada, Toronto; 147p.

Page 40: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

39

OBJECTIVES, INDICATORS, AND REFERENCEPOINTS FOR CONSERVING HABITAT

Robert O’BoyleBedford Institute of OceanographyDartmouth, Nova Scotia

The purpose of the talk was to provide an objectives, indicators and reference point framework toguide the conservation of habitat. As many of the concepts were new to the audience, the talkwould hopefully stimulate development of new ideas and approaches. First, some background onconcepts was presented. The findings of the 2000 ESSIM Workshop were presented, specificallythe main elements of a precautionary approach (objectives, indicators and reference points,account for uncertainty, decision rules and performance measurement) and the key objectives ofecosystem-based management (EBM). The latter include the conservation of biodiversity,productivity and water quality. It was highlighted that the DFO National Policy Committee hadconsidered the ESSIM objectives and struck a national working group to develop them further ata national workshop (Sidney, BC in February 2001). The latter redefined the ESSIM Workshopobjectives by replacing water quality with habitat and by further developing the hierarchicalstructure under each, which was presented. The process (‘unpacking’) whereby operationalobjectives are developed from these conceptual objectives was outlined, along with theterminology used. Examples of unpacking for diversity and habitat were discussed for illustration.It was noted that, similar to harvest fisheries, ‘control rules’ are needed for conservation ofhabitat. This led to a discussion on how multi-indicator arrays for habitat can be assessed using aTraffic Light Approach, which has been used for harvest fisheries. Some of the strengths andweaknesses of the approach were outlined. The talk concluded with a call for an ‘unpacking’workshop on a specific habitat issue to test the concepts and approach.

Page 41: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

40

AVAILABILITY OF INFORMATION ON BENTHICMACROINVERTEBRATE COMMUNITIES FOR CLASSIFYING

MARINE ENVIRONMENTS ON THE SCOTIAN SHELF

Patrick L. Stewart1 and Barry T. Hargrave2

1 Envirosphere Consultants Limited, PO Box 2906, Unit 5—120 Morison Drive, Windsor, NovaScotia B0N 2T0.2 Fisheries and Oceans Canada, Oceans and Environment Branch, Maritimes Region.

The macrobenthic communities (organisms retained on an 0.5 mm sieve) of the Scotian Shelf,the continental shelf off Nova Scotia from Northeast Channel on the Southwest to the LaurentianChannel on the Northeast, are comparatively little studied. No comprehensive studies to samplecommunities and determine their distribution on the shelf as a whole have been carried out;however a range of independent studies both carried out over time and ongoing, though still smallin number, are gradually improving the understanding of this part of the eastern Canadiancontinental shelf. The level of knowledge of macrobenthic communities on the Scotian Shelf isless extensive, however, than for shelf environments in other areas such as the northeasterncontinental shelf of the US (e.g. Wigley and McIntyre, 1964; Wigley, 1962; Emery, et al., 1965;Theroux and Wigley, 1998; Maciolek-Blake, et al., 1985) and on the European continental shelf(e.g. Buchanan, 1963; Duineveld, et al., 1991).

Current knowledge of benthic communities comes from a range of sources. A comprehensivereview of anecdotal surveys and reports of spot sampling pre-1983 is summarized in Mobil (1983)(Figure 1) and more recently in SOEP (1996) and Wildish (1998). Studies have included bottomsampling in support of Russian fisheries investigations which sampled a handful of stations on theeastern Scotian Shelf (Nesis, 1963; 1965) (Figure 2); bottom photographic surveys of geologicalsurface features and distribution on Sable Island Bank and the Northwestern Gully (Stanley andJames,1971); US geological and fisheries surveys extending into Canadian waters (SouthwestScotian Shelf on Browns Bank and vicinity, Emery, et al., 1965; Wigley and McIntyre, 1964;Theroux and Wigley, 1998) research studies on benthic processes in specific locations (BrownsBank, Wildish, et al., 1989; 1993; Mills and Fournier, 1979), transect between Emerald Basin andthe Scotian Shelf (Grant, et al., 1987; 1991) Emerald Bank, Basin and Continental Slope; andVolckaert (1987) St. Margaret’s Bay and Emerald Basin. Recent localized data collection studiesby Fisheries and Oceans Canada in support of strategic objectives (e.g. impacts of trawling,Prena, et al., 1996) and impacts of drilling wastes, as well as other data collected by DFO(biomass data summarized in Stewart, et al., 2001); and studies to provide background for marineprotected area development in The Gully—Hargrave and Hawkins (2001 MS); Kostylev (2001MS) (Figure 3). Recently, biological/geological studies to develop remote mapping technology forbiological community types (Browns Bank, Kostylev, et al., 2001) and to determine distribution ofcommunity types in support of offshore hydrocarbon development on the Scotian Slopesouthwest of Sable Island and off Logan Submarine Canyon (Geological Survey of Canada,2000a and b) have been carried out (Figure 3). Sampling of benthic communities is alsocommonly carried out as part of environmental monitoring activities at offshore wellsites; severalstudies carried out in the 1990s and largely unpublished (summarized in Stewart, et al., 2001),and similar ongoing and new studies will be a source of information, though localized, in futurewhile proposed efforts such as the SEAMAP project may yield information on benthiccommunities through biological samples obtained for ground-truthing the acoustic and bottomphotographic data. In addition, independent and anecdotal studies of distribution of coldwatercorals (Breeze, 1997; Breeze and Davis, 1998) have been carried out; and resource surveys forcommercial clam species on major banks (Rowell and Chaisson, 1983; Chaisson and Rowell,1985; Roddick and Lemon, 1992; Roddick and Kenchington, 1990; Roddick, 1996) have alsoprovided data on components of the benthos (Figure 4). Some of these studies (e.g. Roddick andKenchington, 1990), also included lists of invertebrate by-catch of clam surveys. Resourcesurveys for commercial shrimp (e.g. Koeller, 1996a and b) are also relevant to characterizebenthic communities in certain areas. Biomass data from the Scotian Shelf from all studies to1998 are summarized in Stewart, et al., (2001) and are currently available on the DFO, MarineFish Division Virtual Data Centre, through the DFO Intranet (Figure 5). Part of this project will use

Page 42: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

41

biomass data and its relationship with sediments, depth and other factors to estimate patterns ofbiomass distribution on the Scotian Shelf as a whole. Examination of biomass data for areaswhere both commercial clam surveys and DFO grab sampling in support of trawling impactsstudies has taken place, gives an indication of the size spectra of biomass, and also thecontribution of different groups. DFO commercial clam surveys (e.g. Rowell and Chaisson, 1983;Roddick, 1996) sample the size fraction above 5 cm over large areas of Banquereau and some ofthe other banks, and have given good estimates of biomass in that size fraction over large areasof Banquereau and Sable Island Bank. Grabs taken in the same area as some of the trawl sitesshow a lower biomass, and also different species composition, in the smaller size fraction (Figure6). In grabs, the high biomass is dominated by the Northern Propeller Clam (Cyrtodaria siliqua)compared with the Stimpson’s surf clam (Mactromeris polynyma) in the commercial clam surveys.In some areas, large clam species are responsible for the majority of the biomass (Figure 7). Thistype of information coupled with knowledge of bottom type can be used to provide estimates ofbiomass over wide areas of the shelf.

Several maps of distribution of benthic communities on the Scotian Shelf, based on existing data,have been assembled, in all cases based on limited data, and probably do not represent actualcommunities and dominance relationships very well. More recent studies, such as the samplingdone for assessing the impacts of trawling, although still limited, suggest that there is significantregional variation in communities, which is not reflected in the early representations. Nesis (1965)provided the first distribution map, which covered only the eastern Scotian Shelf in addition to theGrand Banks and Labrador continental shelf (Figure 8).

Steele, et al., (1979) summarized benthic communities on the Scotian Shelf from limited existinginformation (Figure 9), and incorporated the map from Nesis (1965), in an exercise to providesupporting information in Parks Canada’s efforts to identify representative terrestrial and marineareas for the national parks system. A subsequent map of benthic communities was preparedfrom this and other data as part of the environmental impact statement for the Venture gas projectin the mid-1980s (Mobil, 1983) (Figure 10), and was updated for the Sable Offshore EnergyProject environmental impact statement (SOEP, 1996) (not shown). The latter was largely basedon the pre-1983 data presented in Mobil (1983) with the addition of information on distribution oflarge clam species (Mactromeris, Arctica, and Cyrtodaria) obtained from commercial clamsurveys on the Scotian Shelf (Rowell and Chaisson, 1983; Chaisson and Rowell, 1985; etc.) (seeabove) as well as known areas of scallop distribution from commercial scallop surveys carried outby DFO.

The paucity of information on macrobenthic communities makes it difficult to use them inclassifying benthic environments for the Scotian Shelf as a whole, though enough informationmay be available to begin to characterize specific areas. Surficial geology can provide a structureor framework on which the biological communities can be associated, although the conditionsconducive to their development depend on a range of physical and biological conditions, many ofwhich are equally important in determining communities which occur there (Figure 11). Recently,surficial geology maps for the Scotian Shelf have been digitized (MapInfo) by DFO, Oceans andCoastal Management Division, and will provide an additional tool in studying benthic communitiesin the area (Figure 12).

An approach developed by the Geological Survey of Canada, Atlantic Geoscience Centre,utilizing physical characteristics of the bottom determined by multibeam sonar, sidescan, andacoustic reflectance studies, coupled with bottom photographs, calibration by grab samples, andmultivariate analysis (e.g. Kostylev, et al., 2001) is a useful way to characterize large scaledistribution and variations in seabed communities. Within those large scale features it may bepossible to use the geological understanding and knowledge of distribution of bedforms (surfacefeatures of a scale of a few centimetres to kilometres), such as sand ripples, sand waves,megaripples, and sand ridges (see Amos and King, 1984 ) to predict the occurrence of benthiccommunities. An understanding of the dynamics of many of these surface bedforms has beendeveloped for parts of the Scotian Shelf. Sediment type is one of several factors in the biology ofbenthic communities, and influences biology at several scales (Figure 13). Small scale features

Page 43: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

42

having dimensions ranging from centimetres to tens of centimetres include sand ripples causedby waves and currents (Amos, et al., 1988), shell and detritus, all of which are mobile andtemporally unstable, and are probably least significant in determining biological community type.Medium scale features, ranging from tens of centimetres to kilometres, include bedforms such assand waves, sand ribbons, sand ridges, and megaripples (Amos and King, 1984) on which it maybe possible to determine small scale distribution of biological communities (Figure 14). Still largerare the extensive features which can be identified by the approach using multibeam, sidescan,and multivariate approaches as noted above for Browns Bank, which are subunits of the zones ofsimilar sediment types and origins presented on the surficial geology maps for the area (e.g.King, 1970) which represent the highest level. The bottom types identified by the multibeam sonarand multivariate methods are likely to be the most useful in overall classification of the shelf formanagement purposes, while the small features within these types are the ‘texture’ or‘environmental grain’, to which small scale distributions of biological communities can becorrelated, and extrapolated to the larger context, providing important input on the quality andimportance of these bottom types.

Several examples of repeating geological features which have distinctive biological attributes areavailable. Bioherms (biological communities associated with particular seabed geology features)have been studied in the Bay of Fundy; beds of horse mussel of (Modiolus modiolus) areassociated with repeating ridges (Wildish, et al., 1998) (Figure 15). Benthic productivity of thesesites has been estimated and can provide an overall ranking of the importance of these featuresto other environments. Similarly, in the English Channel, bottom communities have been linked totidally generated seabed sand ribbons and furrows (Holme and Wilson, 1985) (Figure 16). In asimilar way, biological communities can be shown to vary with location in relation to position onsand ridges and trough systems off Sable Island. Sand ridges have a grain size variation withmedium sand on the tops, fine sand on the down-current slope and coarse sand on the up-current side and trough (Amos and Nadeau, 1988) (Figure 17). Sand ridges are a conspicuousfeature of the seabed south and southeast of Sable Island (Figure 18). Sampling on sand ridgesfor environmental monitoring of offshore hydrocarbon development has demonstrated differentcommunities in the ridges than in the troughs, corresponding to different grades of sediment(Envirosphere Consultants Ltd., 1998) (Figure 19). A similar approach on the Grand Banks atHibernia shows that there are distinct communities on different grades of sand (Hutcheson, et al.,1981). These studies indicate that there is potential for using small scale distribution ofcommunities in relation to geological features to extrapolate distributions to larger areas, usingknowledge of the distribution and relative abundance of bedforms over large areas. In summary,an approach to delineating biological communities based on surficial sediments (as one ofseveral environmental variables) could be to use the large scale mapping capability and existingunderstanding of surficial geology to identify the broadly similar (physically) units; identify theindividual surface feature units or ‘repeating elements’ that make up the ‘texture’ of the units; andconduct small scale biological studies of representative repeating units, which can then be usedto extrapolate to larger areas.

Currently there is not enough information on distribution of seabed biological communities toserve as a basis for classifying seabed environments on the Scotian Shelf. However, we presenthere a number of comments about approaches to the classification of continental shelfenvironments. In the interest of efficiency, we feel it is important to use existing efforts atclassifying shelf environments as much as possible. The process of classifying benthicenvironments is not new and intensive efforts have been made worldwide to develop meaningfulclassification systems (e.g. Ray, et al., 1982; Hayden, et al., 1984; Bailey, 1998), includingvarious efforts to classify marine regions of Canada (e.g. Wiken, et al., 1996; Harper, et al., 1993;Levings, et al., 1996) (Figures 20 and 21). In particular, Parks Canada has commissionednumerous studies to delineate marine regions as part of its efforts to develop its national systemof parks.

One of us was involved in the development of the classification of the offshore regions of theScotian Shelf for the Natural History of Nova Scotia (Davis and Browne, 1997; Davis, et al., 1994)(Figure 22). This exercise involved consideration of physiographic, geological, oceanographic,

Page 44: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

43

and biological features and a review of world classification systems. The approach used in theNatural History of Nova Scotia was modelled, with some modifications, after approaches whichhad successfully been used on terrestrial environments to describe ecologically significant units,but included consideration of the three-dimensional nature and other characteristics of the ocean,such as ocean currents and upwellings. In particular, the physiographic divisions of the ScotianShelf described in King and Fader (1986) were used in subdividing the shelf into useful units. Inaddition, some consideration of existing marine classifications for the Scotian Shelf would ensurethat the system fits both the context of previous knowledge and as well provides a framework intowhich management systems for adjacent regions (e.g. Newfoundland) could fit. Someclassification efforts such as the classification of Large Marine Ecosystems (LMEs) (Sherman, etal., 1990) (Figure 23) may not be suitable, however, having been set up with only limitedconsideration of global classification then available, and are not as appropriate as those such asWiken, et al., (1996) which present extensions of an accepted global system. Efforts on marineclassification on the west coast of Canada (Levings, et al., 1996) have already productively usedthe marine classification presented in Wiken, et al (1996), and we would recommend that at thevery least, the Scotian Shelf should as well, falling in the ‘Atlantic Marine Ecozone’, which is theappropriate zone from that system.

Systems which use a biophysical approach are to be preferred since biological systems on theseabed and in the water column depend ultimately on both physical and biological factors. Forexample, although surficial geology distribution can influence benthic communities, types ofcommunities are also related to water temperature; benthic biomass and productivity arecommonly directly related to water column primary production (Hargrave and Peer, 1973;Figure 24); and benthic biomass is inversely related to depth, reflecting the utilization of carbonfixed in the water column and decreasing amounts reaching the bottom with increasing depth(Figure 25). Except in shallow nearshore waters where seaweed production is equally important,the seabed communities on the continental shelf are dependent on production in the watercolumn.

Benthic processes are closely connected with water column ones, particularly at shallow depthswhere the benthos interacts with water column processes through benthic pelagic coupling,releasing nutrients back into the water column. In places where the mixed layer meets theseabed, water column production can be channelled directly to the seabed through settling ofphytoplankton and detrital particles. It could be hypothesized that the zone above the typicalmixed layer depth might support distinctive biological communities. Objective biophysicalclassifications such as that developed by the World Wildlife Fund (WWF) (Day and Roff, 2000),although at an early stage of development, can identify bottom communities which may besignificant, such as a cold-water dominated zone off Cape Breton Island and the shallow zonearound Sable Island which corresponds to the zone above the approximate depth of the mixedlayer (Figures 26 and 27). Approaches of this type with refinement and input of latest data onbiophysical parameters of temperature, circulation, and depth, as well as biological parameterssuch as productivity, may be able to provide refined classifications based on the WWF‘landscape’ model.

In summary, a classification approach that would be suitable would: 1) be consistent with existingglobal systems and approaches, at least at the largest scale; 2) include the three dimensionalnature of the marine environment; 3) deal with the pelagic and benthic communities of givenareas as parts of the same ecosystem; and 4) be flexible enough to deal with the absence ofknowledge about some ecosystem components.

Current interest in management of Canada’s continental shelves, and concerns over conservationof a range of species—from whales to coldwater corals; recognition of impacts of long-standingindustries such as the offshore trawling fishery; and increasing interest in development of offshorehydrocarbon resources on the Scotian Shelf; as well as development of new geologicaltechniques, have begun to extend our knowledge of and interest in offshore benthic communities.This knowledge will be useful in understanding and classifying the Scotian Shelf for oceanmanagement purposes in the future.

Page 45: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

44

References

Amos, C.L., and E.L. King. 1984. Bedforms of the Canadian eastern seaboard: A comparisonwith global occurrences. Mar. Geol. 57:167-208.

Amos, C.L., and O.C. Nadeau. 1988. Surficial sediments of the outer banks, Scotian Shelf,Canada. Can. J. Earth. Sci. 25:1923-1944.

Amos, C.L., A.J. Bowen, D.A. Huntley, and C.F.M. Lewis. 1988. Ripple generation under thecombined influences of waves and currents on the Canadian continental shelf.Continental Shelf Research 8:1129-1153.

Bailey, R.G. 1998. Ecoregions. The Ecosystem Geography of the Oceans and Continents.Springer-Verlag, New York.

Breeze, H. 1997. Distribution and Status of Deepwater Corals of Nova Scotia. Marine IssuesCommittee. Special Publication No. 1. Ecology Action Centre, Halifax; 61p.

Breeze, H., and D.S. Davis. 1998. Deep Sea Corals. Section 6.5.2; pp. 113-120. In: G. Harrisonand D. Fenton (eds.), The Gully: A Scientific Review of its Environment and Ecosystem.Canadian Stock Assessment Secretariat, Research Document 98/83.

Buchanan, J.B. 1963. The bottom fauna communities and their sediment relationships off thecoast of Northumberland. OIKOS, 14:154-175.

Chaisson, D.R., and T.W. Rowell. 1985. Distribution, abundance, population structure, and meatyield of the ocean quahog (Arctica islandica) and Stimpson's surf clam (Spisulapolynyma) on the Scotian Shelf and Georges Bank. Can. Industry Rept. Fish. Aquat. Sci.155; 125p.

Davis, D.S., and S. Browne. 1997. The Natural History of Nova Scotia. Nova Scotia Museum,Province of Nova Scotia. Nimbus Publishing and Government of Nova Scotia.

Davis, D.S., P.L. Stewart, R.H. Loucks, and S. Browne. 1994. Development of a biophysicalclassification of offshore regions for the Nova Scotia continental shelf. Proceedings ofCoastal Zone ’94, Halifax, Nova Scotia, September 1994.

Day, J.C., and J.C. Roff. 2000. Planning for Representative Marine Protected Areas. AFramework for Canada’s Oceans. Report prepared for World Wildlife Fund Canada,Toronto.

Duineveld, G.C.A., A. Kunitzer, U. Niermann, P.A.W.J. DeWilde, and J.S. Gray. 1991. TheMacrobenthos of the North Sea. Netherlands Journal of Sea Research. 28:53-65.

Emery, K.O., A.S. Merrill, and J.V.A. Trumbull. 1965. Geology and biology of the sea floor asdeduced from simultaneous photographs and samples. Limnol. Oceanogr. 10:1-21.

Envirosphere Consultants Limited. 1998. MS. Sable Offshore Energy Project (SOEP)Environmental Effects Monitoring (EEM), Thebaud Baseline Data Collection CapeMugford Hazards Survey Cruise-- May 1997. Part B: Biological communities. Report toSable Offshore Energy Project, February 1998.

Geological Survey of Canada (Atlantic). 2000a. Cruise Report, Hudson 2000-036. SeniorScientist: DJW Piper. August 2000.

Page 46: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

45

Geological Survey of Canada (Atlantic). 2000b. Cruise Report, Hudson 2000-042. SeniorScientist: D.C. Mosher. September 2000.

Grant, J., F. Volckaert, and D.L. Roberts-Regan. 1987. Resuspendable Organic Matter in NovaScotian Shelf and Slope Sediments. Continental Shelf Research 7(9):1123-1138.

Grant, J., C.W. Emerson, B.T. Hargrave, and J.L. Shortle. 1991. Benthic oxygen consumption oncontinental shelves off Eastern Canada. Continental Shelf Research 11:1083-1097.

Hargrave, B.T., and C.M. Hawkins. 2001. Patterns of epifauna biomass and respiration in TheGully region of the Scotian Shelf. In: D.C. Gordon and D.G. Fenton (eds.), Advances inUnderstanding The Gully Ecosystem: A Summary of Research Projects Conducted at theBedford Institute of Oceanography (1999-2001). Can. Tech. Rep. Fish. Aquat. Sci.#2377.

Hargrave, B.T., and D.L. Peer. 1973. Comparison of benthic biomass with depth and primaryproduction in some Canadian east coast inshore waters. ICES, C.M. 1973/E:1-14.

Harper, J.R., J. Christian, W.E. Cross, R. Frith, G.F. Searing, and D. Thompson. 1993. Aclassification of the marine regions of Canada. Environment Canada, Ottawa, ON.

Hayden, B.P., G.C. Ray, and R. Dolan. 1984. Classification of coastal and marine environments.Environmental Conservation 11:199-207.

Holme, N.A., and J.B. Wilson. 1985. Faunas associated with longitudinal furrows and sandribbons in a tide-sweptarea in the English Channel. Journal of the marine biologicalAssociation of the U.K. 65:1051-1072.

Hutcheson, M.S., P.L. Stewart, and J.M. Spry. 1981. The Biology of Benthic Communities of theGrand Banks of Newfoundland (including the Hibernia area). Vol. 7. Grand BanksOceanographic Studies. Report for Mobil Oil Canada Ltd., St. John’s, Newfoundland, byMacLaren Plansearch Ltd.; 138p.

Jacques Whitford Environment Limited. 1998. Environmental Effects Monitoring for NorthTriumph. Report to Sable Offshore Energy Inc.

King, L.H. 1970. Surficial geology of the Halifax-Sable Island map area. Marine Sciences Branch,Dept. of Energy, Mines and Resources, Ottawa, Paper No. 1; 16p.

King, L.H., and G.B.J. Fader. 1986. Wisconsinan glaciation of the Atlantic continental shelf ofSoutheast Canada. Bulletin 363, Geological Survey of Canada; 72p.

Koeller, P.A. 1996a. The Scotian Shelf shrimp (Pandalus borealis) fishery in 1995. DFO AtlanticFish. Res. Doc. 96/8.

Koeller, P.A. 1996b. Aspects of the biology of the northern shrimp Pandalus borealis on theScotian Shelf. DFO Atlantic Fish. Res. Doc. 96/9.

Kostylev, V.E. 2001. Benthic assemblages and habitats of the Sable Island Gully. In: D.C. Gordonand D.G. Fenton (eds.), Advances in Understanding The Gully Ecosystem: A Summaryof Research Projects Conducted at the Bedford Institute of Oceanography (1999-2001).Can. Tech. Rep. Fish. Aquat. Sci. #2377.

Kostylev, V.E., B.J. Todd, G.B.J. Fader, R.C. Courtney, G.D.M. Cameron, and P.A. Pickrill. 2001.Benthic habitat mapping on the Scotian Shelf based on multibeam bathymetry, surficialgeology and sea floor photographs. Marine Ecology Progress Series 219:121-137.

Page 47: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

46

Levings, C.D., J.D. Pringle, F. Aitkens (eds.). 1998. Approaches to marine ecosystem delineationin the Strait of Georgia: Proceedings of a DFO Workshop, Sidney, B.C., 4-5 November,1997. Can. Tech. Rep. Fish. Aquat. Sci. 2247; 165p.

Li, M., R. Currie, D. Beaver, P. Girouard, and P. Pledge. 1999. Multibeam Surveys of SableIsland Bank and the Gully Area of Scotian Shelf: Report on Creed Cruise 98-100.Geological Survey of Canada, Internal Cruise Report, March 1999.

Maciolek-Blake, N., J.F. Grassle, and J. Neff (eds.). 1985. Georges Bank Benthic InfaunaMonitoring Program. Final Report of the Third Year of Sampling. Volumes 1-3 andAppendices. Battelle New England Research Laboratory, Duxbury, MA 02332, USA.

Mills, E.L., and R.O. Fournier. 1979. Fish production and the marine ecosystems of the ScotianShelf. Marine Biology, 54:101-108.

Mobil Oil Canada, Ltd. 1983. Venture Development Project Environmental Impact Statement.Volume IIIb Biophysical Assessment; 184p.

Nesis, K.I. 1963. Soviet investigations of the benthos of the Newfoundland-Labrador fishing area;pp. 214-220. In: Y.Y. Marti (ed.), Soviet Fisheries Investigations in the Northwest Atlantic,1963. Dept. of Interior and National Science Foundation.

Nesis, K.I. 1965. Biocoenoses and biomass of benthos of the Newfoundland-Labrador region.Fish. Res. Board Can. Transl. Ser. No. 1375; 74p. (in Russian).

Parsons, J., and Associates. 1994. Cohasset Oil Based Drilling Muds Environmental MonitoringProgram, Lasmo Nova Scotia Limited - 1993 Program Results. Report to LASMO NovaScotia Limited.

Prena, J., T.W. Rowell, P. Schwinghamer, K. Gilkinson, and D.C. Gordon, Jr. 1996. Grand Banksotter trawling impact experiment:1. Site selection process, with a description ofmacrofaunal communities. Can. Tech. Rep. Fish. Aquat. Sci. 2094: viii+38p.

Roddick, D. 1995. Status of the Scotian Shelf Shrimp (Pandalus borealis) Fishery 1994. DFOAtlantic Fisheries Research Document 95/22; 14p.

Roddick, D. 1996. The Arctic surfclam fishery on Banquereau Bank. DFO Atlantic FisheriesResearch Document 96/36; 17p.

Roddick, D.L., and D. Lemon. 1992. Exploratory Survey for Small Arctic Surfclams on the EasternScotian Shelf. Can. Ind. Rep. Fish. Aquat. Sci. 215; 33p.

Roddick, D.L., and E. Kenchington. 1990. A Review of the Banquereau Bank Fishery forMactromeris polynyma for the 1986 to 1989 Period. CAFSAC Research Document 90/14;27p.

Rowell, T.W., and D.R. Chaisson. 1983. Distribution and abundance of the ocean quahaug(Arctica islandica) and Stimpson's surf clam (Spisula polynyma) resource on the ScotianShelf. Can. Industry Rept. Fish. Aquat. Sci. 142; 75p.

Ray, G.C., B.P. Hayden, and R. Dolan. 1982. Development of a biophysical coastal and marineclassification system; pp. 39-45. In: J.A. McNeely and K.R. Miller (eds.), National Parks,Conservation, and Development. The Role of Protected Areas in Sustaining Society.Smithsonian Institution Press, Washington.

Sable Offshore Energy Project (SOEP). 1996. Environmental Impact Statement Volume 3. SableOffshore Energy Project.

Page 48: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

47

Sherman, K., L.M. Alexander, and B.D. Gold (eds.). 1990. Large Marine Ecosystems. Patterns,Processes, and Yields. American Association for the Advancement of Science,Washington,DC.

Stanley, D.J., and N.P. James. 1971. Distribution of Echinarachnius parma (Lamarck) andassociated fauna on Sable Island Bank, southeast Canada. Smithsonian Cont. Earth Sci.No. 6: 24p.

Steele, D.H., J.M. Greene, and J. Carter. 1979. A Biological and Oceanographic Study of theAtlantic Southeast Coast Marine Region. A Report Submitted to Parks Canada,Department of Indian and Northern Affairs, January 1979.

Stewart, P.L., H.A. Levy, and B.T. Hargrave. 2001. Database of Benthic Macrofaunal Biomassand Productivity Measurements for the Eastern Canadian Continental Shelf, Slope andAdjacent Areas. Can. Tech. Rep. Fish. Aquat. Sci. 2336.

Theroux, R.B., and R.L. Wigley. 1998. Quantitative Composition and Distribution of theMacrobenthic Invertebrate Fauna of the Continental Shelf Ecosystems of theNortheastern United States. NOAA Technical Report, National Fisheries Marine Services140. U.S. Department of Commerce.

Volckaert, F. 1987. Spatial patterns of soft-bottom Polychaeta off Nova Scotia, Canada. MarineBiology 93:627-640.

Wigley, R.L. 1962. Benthic Fauna of Georges Bank. Transactions of the North American Wildlifeand Natural Resources Conference 26:310-317.

Wigley, R.L., and A.D. McIntyre. 1964. Some Quantitative Comparisons of Offshore Meiobenthosand Macrobenthos South of Martha’s Vinyard. Limnology and Oceanography 9:485-493.

Wiken, E.B., D. Gauthier, I. Marshall, K. Lawton, and H. Hirvonen. 1996. A Perspective onCanada’s Ecosystems. An Overview of the Terrestrial and Marine Ecozones. CanadianCouncil of Ecological Areas Occasional Paper No. 14.

Wildish, D.J. 1998. Benthos. Section 6.5.1; pp. 107-112. In: G. Harrison and D. Fenton (eds.),The Gully: A Scientific Review of its Environment and Ecosystem. Canadian StockAssessment Secretariat Research Document 98/83.

Wildish, D.J., A.J. Wilson, and B. Frost. 1989. Benthic macrofaunal production of Browns Bank,Northwest Atlantic. Canadian Journal of Fisheries and Aquatic Sciences 46:584-590.

Wildish, D.J., A.J. Wilson, and B. Frost. 1992. Benthic Boundary Layer Macrofauna of BrownsBank, Northwest Atlantic, as Potential Prey of Juvenile Benthic Fish. Canadian Journalof Fisheries and Aquatic Sciences 49:91-98.

Wildish, D.J., G.B.J. Fader, P. Lawton, and A.J. MacDonald. 1998. The acoustic detection andcharacteristics of sublittoral bivalve reefs in the Bay of Fundy. Continental Shelf Research18:105-113.

Page 49: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

48

Figure 1. Summary of sources of information on benthic communities of the Scotian Shelfcarried out pre-1983 (Mobil, 1983).

Page 50: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

49

Figure 2. Studies of Benthic Invertebrate Biomass on the Scotian Shelf, 1965-1998 (fromStewart, et al., 2001).

Figure 3. Recent studies of benthic communities using combined acoustic and geologicalmethods, bottom photographs and spot grab sampling to classify extensive areas ofthe seabed.

Page 51: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

50

Figure 4. Locations of sampling for commercial molluscs, Mactromeris and Arctica on theScotian Shelf (Chaisson and Rowell, 1985).

Figure 5. Mapping capability for macrobenthic biomass data in the DFO, Marine Fish DivisionVirtual Data Centre.

Page 52: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

51

Figure 6. Biomass of the benthic community at a location on Banquereau Bank is greater thanfor the size fraction containing the commercial species (Stimpson's Surf Clam,Mactromeris polynyma). (Data source: Fisheries and Oceans Canada, KevinMacIsaac).

0

500

1000

1500

2000

2500

3000

3500

Wei

ght o

f Tax

on (g

/m2)

0 1000 2000 3000 4000 5000 Total Wet Weight Biomass (g/m2)

Echinoderms Mollusca

BANQUEREAU BANKMollusc & Echinoderm Biomass vsTotal

MolluscRegressiony = 0.98x - 263r2 = 0.919n = 51

Figure 7. Relationship between benthic biomass and biomass of molluscs and echinodrems,Banquereau Bank (Data source: Fisheries and Oceans Canada, Kevin MacIsaac).

Page 53: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

52

Figure 8. Biocoenoses of the Canadian eastern continental shelf and slope (Nesis, 1965).

Figure 9. Bottom communities on the Canadian eastern continental shelf and slope (Steele, etal., 1979).

Page 54: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

53

Figure 10. Characteristic benthic communities from the central Scotian Shelf (Mobil, 1983).

Determinants of Benthic Biological Communites

> geography/water masses/temperature> food supply> bottom substrate and stability> interactions with other species (e.g. predation)

Figure 11. Determinants of benthic biological communities.

Page 55: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

54

Figure 12. Distribution of surficial sediments, Scotian Shelf.

Scales of Seabed SedimentVariability

• centimetres to tens of centimetres(sand ripples, rocks, stones shells, territory of

commensal species)• metres to tens and hundreds of metres

(sand waves, megaripples, sand ridges; shell beds)• hundreds of metres to kilometres (sand ridges, gravel waves)• kilometres to tens and hundreds of kilometres

(geological scale sediment facies)

Figure 13. Scales of seabed sediment variability.

Figure 14. Surficial sediment bedforms on the eastern Canadian continental shelf (Amos andKing, 1984).

Page 56: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

55

Figure 15. Bioherms (seabed features with associated biological communities) in the Bay ofFundy (Wildish, et al., 1998).

Figure 16. Biological communities are associated with seabed geological features in the EnglishChannel (Holme and Wilson, 1985).

Page 57: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

56

Figure 17. Sediment grain size variation across a sand ridge, Banquereau Bank (Amos andNadeau, 1988).

Page 58: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

57

Figure 18. Sand ridge formations south of Sable Island in the vicinity of the Thebaud well site (Li,et al., 1999 MS). Dot indicates location of wellsite.

Page 59: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

58

Figure 19. Multidimensional scaling plot of stations on sand ridge system based on benthiccommunity composition. (Envirosphere Consultants Limited, 1998).

Page 60: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

59

Figure 20. Ecoregions of Canada (from Wiken, et al., 1996).

Hierarchy of Ecosystems• eco-zones• eco-regions• eco-districts• eco-sections• eco-sites• eco-elements

Figure 21. Ecosystem regions for the classification of terrestrial and marine regions of Canada(Wiken, et al., 1996).

Page 61: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

60

Figure 22. Classification of offshore regions from the Natural History of Nova Scotia (Davis andBrowne, 1997).

Page 62: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

61

Figure 23. Large marine ecosystems (Sherman, et al., 1990).

Page 63: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

62

Figure 24. Relationship of benthic community biomass to primary productivity in the watercolumn (from Hargrave and Peer, 1973).

0

100

200

300

400

500

Biom

ass

(g/m

2 w

et w

eigh

t)

0 200 400 600 800 1000 Depth (m)

Figure 25. Typical variation of benthic community biomass with depth between continental shelfand slope environments (from Stewart, 1983; for the continental shelf of the DavisStrait and Ungava Bay).

Page 64: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

63

Figure 26. Location of the 50m contour on Sable Island Bank, as an indication of the contact ofthe surface mixed layer with the seabed.

Page 65: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

64

Figure 27. Classification of environments on the Scotian Shelf based on a 'landscape' approach(Day and Roff, 2000) is an objective approach which identifies regions which may bebiologically significant.

Page 66: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

65

EUNIS CLASSIFICATION: AN EXCELLENT EXAMPLE OFA WELL-DEVELOPED GENERAL CLASSIFICATION

SYSTEM FOR MULTIPLE USES

Paul BoudreauMarine Environmental Sciences DivisionDepartment of Fisheries and OceansBedford Institute of OceanographyDartmouth, Nova Scotiahttp://www.mar.dfo-mpo.gc.ca/oceans/e/mesd/mesd-e.htmlhttp://www.bio.gc.ca/

The European Union Nature Information System (EUNIS) has been under development for anumber of years with the goal of establishing a useful habitat classification system for allterrestrial and aquatic lands.

The goals of the effort are to:

• provide a “common language”;• enable mapping of units at a regional level;• be comprehensive and applicable at different levels of complexity;• allow aggregation, evaluation and monitoring of habitat units; and• provide a common framework for new information and links to other classifications.

It is based on a number of previous classification efforts and builds on them by using a number ofworkshops and consultations with researchers with experience in the various aquatic habitats.The International Council for the Exploration of the Seas (ICES) has played a role in this processand has accepted the EUNIS classification down to level 3 as a working framework forimplementing and testing its usefulness.

The principles of the EUNIS classification are:

• Classification is hierarchical;• Units at a given hierarchical level to be of similar importance;• Clear criteria for each division;• Logical sequence of units;• Use clearly defined non-technical language;• Ecologically distinct habitat types supporting different plant and animal communities should

be separated;• Habitats from different locations differing on the basis of geographical range only should not

be separated; and,• Habitat units and habitat complexes are separated.

These have been worked out through the numerous attempts to apply the classification and arenot always found in other systems.

EUNIS uses a well-documented decision key to separate dissimilar habitats. The high levels aredetermined primarily by water depth and surficial geology/sediment type. The following figureshows the decision system down to level 2 in the classification (Figure 1).

Although ICES has only accepted down to Level 3, the system readily supports the developmentand definition of the additional levels of detail. In the United Kingdom, EUNIS levels 4, 5 and 6have been identified for many habitats. At this higher resolution, biological parameters, such asfunctional groups, species groups and specific indicator species is used.

Page 67: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

66

Each classification and decision point is well defined and supported by documentation. Theextensive web page provides all of the background documentation, including an extensive list ofclassifications:

http://mrw.wallonie.be/dgrne/sibw/EUNIS/eunis.fulllistA.html.

As an ICES country, Canada has committed to considering the application of the EUNIS system,down to level 3, for use in classifying aquatic habitats. Application of this classification in theNorthwest Atlantic will require much additional work to map the necessary underlying bathymetry,sediments and distribution of biological communities. Recognising the limited resources availableto classify and map benthic habitats, the EUNIS system may be worth considering as a provenworkable framework, even if it may not be totally transferable.

Page 68: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

67

No

water column

bedStratum

(17)

AMARINE

HABITATS

AMARINE

HABITATS

A7Pelagic water

column

A7Pelagic water

columnSubstrate

(21)

No

soft or mobilesediments

hard or non-mobilesubstrates

Substrate (19)

Depth>200m?

(20)

Yes

Permanentlywater-covered?

(18)

soft or mobilesediments

hard or non-mobilesubstrates

A5Bathyal zone

A5Bathyal zone

Yes

Slope and rise? (22)

A6Abyssal zone

A6Abyssal zone

NoYesA4

Sublittoralsediments

A4Sublittoralsediments

A3Sublittoral rock andother hard substrata

A3Sublittoral rock andother hard substrata

A2Littoral sediments

A2Littoral sediments

A1Littoral rock and

other hard substrata

A1Littoral rock and

other hard substrata

Figure 1. Decision tree down to level 2 in EUNIS habitat classification system.

Page 69: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

68

THE SURFICIAL SEDIMENTS OF THE SCOTIAN SHELF:A REVIEW OF PUBLISHED MAPS, LIMITATIONS, FUTURE

USES AND RECENT ADVANCES IS SEABED MAPPING

Gordon B.J. FaderGeological Survey of Canada (Atlantic)Bedford Institute of OceanographyDartmouth, Nova Scotia

The surficial sediments on the Scotian Shelf were studied and mapped by the Geological Surveyof Canada in a first round of seabed mapping that occurred from 1967 to 1984. The program wasregional in scope and began with the production of the first map in an area extending from Halifaxto Sable Island (King, 1970) where the methodology was developed. Subsequent studies andmaps in the series by Fader, et al., (1977), Drapeau and King (1972), MacLean and King (1971),MacLean, et al., (1977), Fader, et al., (1988), and Fader, et al., (1982), produced anunderstanding of the surficial geology of the Scotian Shelf and adjacent areas. Kranck (1971)followed with a similar approach for areas of the Northumberland Strait, and Loring and Nota(1973) mapped the surficial geology of the Gulf of St. Lawrence with an emphasis on sampleanalysis. From these studies, a formal stratigraphy for the Scotian Shelf was defined and refined(King and Fader, 1986). Summary surficial sediment compilations were produced for the ScotianShelf Basis Atlas Series (Fader, 1991). All of the surficial geological maps are accompanied bydetailed reports on methodology with details on sediment distribution, character and stratigraphy.

The maps produced from this systematic mapping program were based on interpretation ofechograms of the seabed, which are similar to high-resolution seismic reflection profiles,combined with analysis of many thousands of grab samples and bottom photographs.Interpretation of echograms considered acoustic penetration, seabed hardness, internal unitreflection character, seabed micro and macro roughness, and sediment stratigraphy. Fivesurficial sediment units, termed formations, were identified and mapped (Scotian Shelf Drift,Emerald Silt, Sambro Sand, Sable Island Sand and Gravel and LaHave Clay), and a chronologyof the geological history was proposed. The dominant controlling factors on sediment distributionand character were : 1) shelf wide glaciation which resulted in erosion and deposition of glacialmaterials, 2) lowering of sea levels to 110m below present sea level, and 3) a final marinetransgression of advancing beach processes to the present coastline. These are the keyconcepts that lie behind the interpretation of the sediments and the production of the maps, whichdisplay a mix of fact and interpretation. The maps and associated reports have provided aconceptual framework for knowledge of the distribution and character of sediments on the ScotianShelf. Over 60% of the sediments on the Scotian Shelf can be described as being predominantlyrelict, that is, exhibiting characteristics of past and non-active environments, with little modernmodification. Most recent maps in the series benefited from advancing technology such as theapplication of high-resolution seismic reflection profilers and sidescan sonar systems (Fader, etal., 1982).

Recently there has been a renewed interest in the surficial maps of the Scotian Shelf in responseto Canadian requirements for improved management of the seabed derived from legislation in theOceans Act that includes diverse activities such as habitat characterization, conflict resolution,resource extraction and conservation, and the selection of Marine Protected Areas. The questionthat arises in light of these requirements is to the validity and utility of the existing maps for suchapplications.

For proper use of the seabed to be effected and conflicts to be avoided or mitigated, detailedknowledge of living and non-living seabed resources, as well as seabed characteristics andprocesses, is essential. Users of the seabed generally require three categories of knowledge.

Page 70: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

69

These are:

1. the distribution of materials and their associated properties;2. the morphology of the seabed (commonly defined as bathymetry); and3. an understanding of seabed dynamics (erosion, deposition, sediment transport, stability).

To evaluate the existing maps from this perspective it is necessary to assess how they addresseach of the three requirements. The most important contribution of the maps is their portrayal ofsediment type, classification and distribution in a conceptual model framework. This was theprime intention of the initial mapping program. The formational basis of sediment classificationenables complex sediment variations to be combined in a logical, simplified and workablegeological framework that can accommodate sampling problems and errors, minor localvariations, post-depositional winnowing and a lack of continuous seabed information. Simpletextural maps based on sample control alone would not be able to characterize the seabed in acoherent geological sense and would be of little practical value. For example, the classification ofglacial till (Scotian Shelf Drift) allows a wide variety of mud, sand and gravel mixtures to besummarized within a coherent depositional unit directly deposited by glacial ice as a moraine.

On the other hand, the morphology of the seabed is poorly represented on the surficial maps. Thebathymetry was extracted from published hydrographic charts produced at similar scales andprojections and intended for safe navigational purposes and not intended (or able) to portray thedetailed morphology of the seabed. In many areas the survey control is spaced at distances ofhundreds of meters to several kilometres, thus small-scale morphological variations over featuressuch as sand waves, iceberg furrows, pockmarks, and bedrock are not resolved.

Regarding a dynamic assessment of the seabed, the maps also poorly represent conditions ofsediment transport and deposition. In a broad sense, depositional areas can be defined where,for example, LaHave Clay deposits occur, but details of dynamics are not portrayed nor are theyinterpretable. Sidescan sonar data, which clearly portrays dynamic sediment features, was largelyunavailable for the production of the earlier maps. However, later maps on the eastern andwestern areas of the Scotian Shelf depict fields of bedforms (sand waves) from interpretation oflimited sidescan sonar coverage.

This assessment clearly shows that the earlier maps, although rich in ground-truthed sedimentdata, are lacking in morphological detail and dynamic content. The recent application of newmultibeam bathymetric mapping techniques to several previously mapped areas has given veryhigh-resolution (decimetre) insight into morphological character and dynamic processes active onthe continental shelf. Such dynamic processes, interpretable from multibeam data, includeseabed current and wave scouring, non-depositional moat development, bedform formation,sediment transport pathways, sedimentary furrow formation, etc. Through digital processingtechniques, multibeam bathymetric data can also be displayed to enhance subtle morphologicalattributes giving considerable insight into previously unknown relationships between currents,waves and resulting seabed processes of erosion and deposition.

The conclusion to be drawn from this assessment is that the existing surficial maps on theScotian Shelf are very limited in their potential to be used for adequate management of seabedrelated activities such as outlined above. The horizontal resolution of the maps, density of controland the precision and accuracy do not meet requirements of many user groups that requiredetailed seabed information. They do, however, provide an essential framework of understandingand can serve as the basis for definition of future detailed study areas and geologicalinterpretation.

In order to address modern issues on the continental shelf involving the seabed, multibeambathymetric mapping and its associated thematic products of morphology, backscatter (proxy forsediment type), slope, interpreted geology and habitat maps are required. Such an approach iscaptured in the Canadian SeaMap proposal to map the offshore areas of Canada with a longterm, systematic program.

Page 71: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

70

If the existing surficial geological maps are to be used in the future until supplanted by new high-resolution multibeam derived products, then there are limitations on their use that must beunderstood by geologists, biologists, managers and policy formulators. For example, it is notadvisable to try and extract site specific textural information from the maps unless the locationrepresents a sample site. The maps were based on textural data from many samples and theactual data base for the site sediment information can be accessed in a GSC Open File Report# 1430 (Sonnichsen, et al., 1987). The coarser the sediment, the less reliable the texturalinformation. This is partially a function of the equipment used to collect samples of the seabedand a general inability to properly sample coarse gravely sediment. Gravel clasts often preventthe jaws of sampling equipment from completely closing which results in a loss of finer-grainedmaterial and a skewness toward coarseness. Thin gravel lag surfaces overlying differentsediment types result in unique sampling problems whereby the sediments often get mixed duringthe sampling process. Therefore, the sediment analysis from coarse sediment (till and gravels) islargely sampler dependent and historically less accurate. Modern seabed sampling with largeseabed invasive samplers, such as the IKU bucket grab, is designed to take large volume grabsto penetrate the sediments and preserve structure down to 0.75m depth.

Fine-grained sediment boundaries, for example, between LaHave Clay and till on the early maps,are generally very accurate, whereas the coarser grained sediment boundaries (between SableIsland Sand and Gravel) are less accurate. This is because the fine-grained clays and silts areeasily differentiated acoustically from coarse sediments and their boundaries can be mapped ingreat detail. Coarse-grained sediments are not easily differentiated with echograms and mappershad to rely on the collection of many samples or other parameters such as roughness andbedforms to differentiate sands and gravels, particularly on the offshore banks.

The early surficial maps also have shortcomings in the nearshore on the inner Scotian Shelf.Here the seabed was mapped as dominantly gravel with minor sand located in isolated patchesand channels. Modern multibeam bathymetry and sidescan sonar systems show large expansesof exposed bedrock, often in ridges in these areas. However, attempts at sampling these surfacesretrieved gravel, which had accumulated at the base of steep bedrock slopes where samplerstended to fall. On echograms, gravel and bedrock have similar acoustic characteristics.

In conclusion, caution should be exercised in using the existing surficial maps for management ofthe seabed of the Scotian Shelf. The Geological Survey of Canada, in a 1996 report on the statusof marine geoscience in Canada, suggested that present knowledge of the vast offshore area isroughly equivalent to what was known about onshore Canada in the late 1800s. Modern mappingtechnologies must be applied to the offshore to adequately address an emerging suite ofmanagement issues.

References

Drapeau, G., and L.H. King. 1972. Surficial geology of the Yarmouth to Browns Bank map area.Marine Sciences, Paper 2, Geological Survey of Canada Paper 72-24; 6p.

Fader, G.B.J. 1991. Surficial geology and physical properties 5: Surficial formations; in EastCoast Basin Atlas Series: Scotain Shelf; Geological Survey of Canada (Atlantic); 119p.

Fader, G.B.J., L.H. King, and B. MacLean. 1977: Surficial geology of the eastern Gulf of Maineand Bay of Fundy; Geological Survey of Canada, Paper 76-17; 23p.

Fader, G.B.J., E. King, R. Gillespie, and L.H. King. 1988. Surficial Geology of Georges Bank,Browns Bank and Southeastern Gulf of Maine; Geological Survey of Canada Open FileReport #1692; 3 maps.

Page 72: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

71

Fader, G.B.J., L.H. King, and H.W. Josenhans. 1982. Surficial geology of the Laurentian Channeland the Western Grand Banks of Newfoundland; Marine Sciences Paper 21; GeologicalSurvey of Canada Paper 81-22; 37p.

King, L.H. 1970. Surficial geology of the Halifax-Sable Island Map Area. Marine Sciences PaperNo. 1; 16p.

King, L.H., and G.B.J. Fader. 1986. Wisconsinan glaciation of the southeastern CanadianContinental Shelf. Geological Survey of Canada, Bulletin 363.

Kranck, K. 1971. Surficial geology of Northumberland Strait. Marine Sciences Branch, Dept. ofEnergy Mines and Resources, Ottawa, Paper No. 5.

Loring, D.H., and D.J.G. Nota. 1973. Morphology and sediments of the Gulf of St. Lawrence.Bulletin 182 of the Fisheries Research Board of Canada.

MacLean, B., and L.H. King. 1971. Surficial geology of the Banquereau and Misaine Bank maparea. Marine Sciences Paper 3, Geological Survey of Canada Paper 71-52; 19p.

MacLean, B., G.B.J. Fader, and L.H. King. 1977. Surficial geology of Canso Bank and adjacentareas. Marine Sciences Paper 20, Geological Survey of Canada Paper 76-15; 11p.

Sonnichsen, G.V., G.B.J. Fader, and R.O. Miller. 1987. Compilation of seabed sample data fromthe Scotian Shelf, The Western Grand Banks of Newfoundland and adjacent areas.Geological Survey of Canada Open File report No. 1430; 328p.

Page 73: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

72

CHALLENGES IN HABITATCLASSIFICATION AND MAPPING

Vladimir E. KostylevNatural Resources CanadaBedford Institute of OceanographyDartmouth, Nova Scotia

Characterization and classification of the benthic environment is required for educatedmanagement of natural resources in Canadian territorial waters. This activity demandsdevelopment of an informed view on the distribution of marine ecosystems, their function, andbiological diversity. It also necessitates geographically accurate representation of acquiredknowledge on maps, which would allow planning of management activities and even navigation.Thus, ultimately, a manager would require a map representing areas, which are ecologicallydifferent and carry meaningful information required for decision making.

Classification vs. Mapping

Two steps in this process are obviously related and interdependent – classification and mapping.While the first does not necessarily require visual representation, the map is a visual aid bydefinition, and while imprecisions in classification systems are hidden behind the terms, the mapsare assumed to correctly represent nature and the user will immediately discover discrepancies inpractice if the map is wrong. Therefore it is important to see the fundamental differences inhabitat classification and habitat mapping – their assumptions, use and implications of possibleimprecision and inaccuracy.

Any verbal classification is to some degree an exercise in coining terms. Ecological classificationsare aimed at pointing out differences between different systems (e.g. shallow vs. deep-waterenvironment), but the borderline between the two is not well defined quantitatively. Because ofthat an arbitrary classification system displayed on a map would not aid the management system,but to the contrary – obscure the patterns being sought.

An example of mapping a classification system is presented as Theme Regions for the ScotianShelf (Davis, et al., 1994). While greatly aiding understanding of the nature of shelf environmenton a large scale, the system is based on a number of set isolines, which serve as guides fordefining the regions, e.g. fishing banks. The portrayal of sea-floor morphology by bathymetriccontours involves a prejudicial simplification of the shapes (Froidefond and Berthois, 1983) andthe confusion is obvious when a person used to metric system tries to find familiar shapes on abathymetric map drawn in imperial units. This is not the only problem. If the shelf is subdividedalong the contours of e.g. 100 and 200-meter isobaths then it is valid to ask if these particularisobats have any ecological significance?

The most notorious bathymetric contour in marine biology is 200 meters. It is commonly used inreference to “deep-sea” environment, lying deeper than the depth of continental shelves, which isassumed to correspond to 200 meters. Francis P. Shepard (1959) in his description of geology ofcontinental shelves writes the following:

“In virtually annexing the shelf, we defined it as the shallow-water area extendingto a depth of 100 fathoms (182.88m). This, however, is purely arbitrary [textomitted] and it is only rarely that they [shelves] terminate at or even close to 100fathoms”.

Joel Hedgpeth (1957) whose bathymetric classification system is referred to in many marinebiology textbooks also mentions the lack of precision in defining different depth zones in the sea:

Page 74: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

73

“It should be understood that we are not referring here to the realities or limits ofthese subdivisions or the problems of zonation on the shore, but to the manner inwhich the terms have been applied”.

The titular nature of classification nomenclature is even more evident from Hedgpeth’s discussionof the use of term “bathyal” that has been applied to the environment of continental slope down toabout 1000-2000 meters.

“We suspect some difference between the environment of the intermediatebottoms and the great abyssal depths, but do not know where transition is”.

There is no controversy that shelf and slope physical environments are considerably different andit is valid to assume that they differ ecologically as well. Current high-resolution bathymetricmapping permits accurate determination of the shelf break and thus, instead of mapping anarbitrary isobath one can draw confident contours of the shelf.

Apparently visualization of an arbitrary classification system immediately uncovers its limitations,and just a minor change in classification system (e.g. replacing 200m depth with seemingly non-quantitative term ‘shelf break’), may lead to ecologically justifiable and scientifically meaningfulmaps that can be used for management purposes. The map is an ultimate test for a classificationsystem and an ultimate product required by managers at the same time.

Top-down vs. Bottom-up Approaches

When used for the purpose of mapping, a-priori classification of environment may be thought ofas ‘top-down’ approach. Assumptions based on the knowledge acquired elsewhere arepropagated along the route from general theory to small-scale spatial definition. The mappingthen is based on the assumption of some relationship between physical factors and biologicalcomponents and thus is strongly based on the combination of maps of physical variables.

‘Bottom-up’ approach leads from observation to generalizations, and is based on grouping ofobservations on the basis of their similarity. Observations are usually well defined spatially andthe errors in defining boundaries of groups of similar observations depend on the intensity andgrain of sampling. The main assumption of this approach is that our sampling correctly representsthe reality, which is not necessarily true. Of course we need a sufficient number of samples tocover the study area.

Habitat maps were created for separate commercial species and have had distinct value forcommercial fishery and management. An ecosystem approach requires knowledge of distributionof spatially defined areas populated by distinct biological communities. Therefore we look atgroups of species and agglomerate similar areas occupied by similar benthic communities.Community ecology, however, has endured two opposite views on community - a ‘superorganism’theory (Clements, 1916) which emphasizes binding between species through their co-evolutionary history, and the view that communities are simply assemblages of species which aredistributed along environmental gradients and co-occur because of similar preferences to theirenvironment (Gleason, 1926). In either case groups of organisms are related to physical factorsand can be used to define sets of physical variables that are important to them.

The current, generally accepted view is closer to the latter, therefore it seems reasonable that it ispossible to start from the factors and proceed ‘down’ to communities. Recall, however, that it isthe ecological niche which is commonly defined as the limits, for all important environmentalfeatures, within which individuals of a species can survive, grow and reproduce (Begon, et al.,1996). Thus the ‘top-down’ approach would map niches for different associations of species, andnot the distinct habitats. This approach disregards the fact that the presence of communitiesdepends not only on a suitability of physical factors, but on the history of ecological succession,colonization and disturbance.

Page 75: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

74

Therefore, the apparent simplicity of a top-down approach is masking two problems – the first oneis the contradiction between the inherent belief of Gleassonian approach in the absence ofcommunities, with defining community boundaries, and the second – mapping of ecologicalniches instead of habitats. While niches are theoretical generalizations, habitats are real bydefinition.

The basic definition of habitat is the following: it is simply the place where a plant or an animallives (Begon, et al., 1996). Unfortunately the term “habitat” may be successfully removed fromtext or replaced by a term “location” in the greater part of scientific literature. Any classification ofhabitats is arbitrary unless it deals with living organisms or their assemblages, and there would beno need for benthic habitat mapping if the Ocean was devoid of life. Habitat of an organism is atangible and spatially defined area where the organism lives, that also implies a set of associatedenvironmental descriptors (e.g. substrate, temperature, salinity, etc.). The characterization ofhabitat requires definition of spatial boundaries and ranges of physical factors based ondistribution of a particular organism or group of organisms that share environmental preferencesand occupy the same locality. Therefore for the purpose of classification and mapping ourworking definition of a habitat should be “a spatially defined area where the physical, chemicaland biological environment is distinctly different from the surrounding”. This definition allowsavoiding differences in theoretical connotation of ecological communities, and permits the use ofstatistical ordination techniques for differentiating distinct benthic environments through theanalysis of distribution of benthic assemblages.

Bottom-up Approach

Mapping habitats from species-up assumes that distribution of animal assemblages adequatelyrepresents environmental factors, which are responsible for shaping modern day communities. Itis assumed that animal communities are indicative of the state of environment during a certainperiod of time and this way data, especially on distribution of sessile species, explicitlyincorporates history of physical environment and existing environmental impacts. Success ininterpretation and interpolation of empirical observations on a map depends, among other things,on spatial heterogeneity of studied system and on the grain of sampling. Computer-generatedisolines are aesthetically appealing and may produce generalized patterns of studied variablese.g. biodiversity. However, the representation may become dubious when more information aboutthe distribution of sampling effort is available and may become completely misleading whenknowledge of the general environmental settings is available. For example one can notinterpolate biomass of sand dollars across a canyon. In the Sable Island Gully (Kostylev, 2001)sampling of different geomorphologies showed the distinct differences in species richnessbetween coarse grained glacially modified terrain and other types of substrates (e.g. bedforms,bank top sands, inner canyon environment). This suggests that mechanistic interpolation andextrapolation of sampling data is meaningless, unless the factors responsible for variability inobservations are revealed and accounted for.

Because mapping is often based on insufficient data, and involves interpolations andextrapolations often rooted in many assumptions, habitat mapping should be consideredgeographically referenced modelling, which should be scientifically defendable, and founded onstrong ecological theory and meticulous observations.

Some Sources of Mapping Error

It is generally agreed that the diversity in habitat types is important for sustaining high biologicaldiversity. It is necessary therefore to have an accurate description of spatial distribution ofdifferent habitats. In defining Marine Protected Areas, for example, it is suggested to assign aregion, that contains the highest diversity of habitat types, adequately represents each of thepresent habitats and has connections to surrounding seascapes as MPA (Day and Roff, 2000). Itis evident that definition of such area requires explicit knowledge of distribution of habitat typesand accurate representation of their boundaries. Boundaries on geographical maps however, areusually of two types – arbitrary and approximate.

Page 76: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

75

Top-down habitat mapping approaches (e.g. Day and Roff, 2000) involve discrimination of habitattypes based on overlaying isolines of different factors and treating their intersections asbiologically meaningful regions. For example overlaying depth, temperature, slope, waterstratification and sediment type isolines produces a multitude of polygons assumed to be distinctenvironments. The rationale behind this is the following: if values of isolines correspond tobiologically meaningful boundary values of relevant physical factors, then each polygon willrepresent a distinct environment, which may have distinctive fauna.

There are several problems with such an approach. First – as discussed earlier some isolines arearbitrary; secondly, the errors in spatial allocation of isolines make intersections meaningless. Asa result, a count of number of polygons within certain area will have a vague if any relation to thediversity of habitats or fauna. There is a multitude of sources of errors, which lead to imprecisionin mapping of individual factors. These are due to natural variability, scarcity of observations,inherent variance in modelling outputs, just to mention a few and just pertaining to variablesdescribing water masses.

Suggestions for Habitat Mapping

Based on our knowledge of benthic communities and their association with the physicalenvironment we can map benthic habitats, being aware of our limitations. It is necessary torecognise that there is a high degree of error in top-down mapping of habitats, and there are veryscarce observations on benthic fauna that are available for a consistent analysis and useful forhabitat mapping. Therefore the only practical approach for broad scale habitat mapping onScotian shelf at this moment would be finding a middle way between top-down and botttom-upapproaches through cross validation. It is necessary to develop statistically and logically validapproaches for dealing with empirical observations and incorporating them into larger picture, i.e.shelf-wide habitat map.

It is often better to see than to assume. Scientific methodology is plagued with assumptions,which are easily overlooked during design, analysis and interpretation of scientific data. Habitatmaps in particular are hypotheses that have to be tested. It is necessary therefore to increasesampling effort on Scotian shelf, through series of dedicated habitat mapping cruises, whichinvolve both DFO and NRCan scientists.

Geological maps offer the highest precision compared to other factors, convey information aboutbathymetry and texture, and habitat complexity, and undoubtedly should be used as a matrix forhabitat mapping.

References

Begon, M., J.L. Harper, and C.R. Townsend. 1996. Ecology: individuals, populations andcommunities. Blackwell Science Ltd., Oxford.

Clements, F.E. 1916. Plant succession: analysis of the development of vegetation. CarnegieInstitute of Washington, Publication No. 242, Washington, DC.

Davis, D.S., P.L. Stewart, R.H. Loucks, and S. Browne. 1994. Development of a biophysicalclassification of offshore regions for the Nova Scotia continental shelf. Proc. CoastalZone Canada Conf., Halifax, NS:2149-2157.

Day, J., and J.C. Roff. 2000. Planning for representative marine protected areas: A framework forCanada’s oceans. World Wildlife Fund Canada, Toronto; 147p.

Froidefond, J.M., and L. Berthois. 1983. Test for the planimetric representation of morphometricparameters adjusted to the study of sea floor features, Bull. Inst. Geol. Bass. Aquitaine.33:39-60.

Page 77: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

76

Gleason, H.A. 1926. The individualistic concept of the plant association. Toorrey Botanical ClubBulletin 53:7-26.

Hedgpeth, J.W. 1957. Classification of marine environments. Geol. Soc. America Memoir 67,1:17-28.

Kostylev, V.E. 2001. Benthic assemblages and habitats of the Sable Island Gully. In: D.C. Gordonand D.G. Fenton (eds.), Advances in Understanding The Gully Ecosystem: A Summaryof Research Projects Conducted at the Bedford Institute of Oceanography (1999-2001).Can. Tech. Rep. Fish. Aquat. Sci. #2377.

Shepard, F.P. 1959. The Earth beneath the Sea. London, Oxford University Press.

Page 78: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

77

HABITAT CLASSIFICATION:A U.S. (NOAA FISHERIES) PERSPECTIVE

Stephen K. Brown, Ph.D.National Oceanic and Atmospheric Administration FisheriesSilver Springs, MarylandUSA

The USA’s National Oceanic and Atmospheric Administration (NOAA) has numerous mandatesinvolving marine habitat (Table 1). The relevant habitats range from the headwaters of salmonspawning streams through the entire U.S. Exclusive Economic Zone, and from the arctic to thetropics. These mandates are fairly recent in origin, so that, in contrast to a more established field,such as fisheries management and stock assessment, NOAA’s basic terminology, concepts, anddata streams are still being defined.

The Sustainable Fisheries Act established new Essential Fish Habitat (EFH) requirements in1996. Fishery management plans are now required to include amendments to identify, conserve,protect, and restore EFH. This requirement has greatly increased the involvement of NOAAFisheries in habitat science and management. At present, many, but not all, of the approximately40 fishery management plans have approved EFH amendments, but, even for the approvedamendments, there are many acknowledged shortcomings in the level of available information.

Some Marine Habitat Classification Systems

Establishment of a widely accepted system of habitat classification is a key aspect of managingand studying habitat. Many systems have been proposed for the marine environment, both in theUSA and in other regions. However, there is no comprehensive or widely accepted classificationsystem.

Habitat classification systems share several characteristics:

• they have a specific purpose or application;• they are applied to some spatial domain and scale;• they are applied to a particular group of organisms; and• they are usually hierarchical.

Some major marine habitat classification systems developed in the U.S. include those ofCowardin, et al., (1979); NOAA Fisheries and the Ecological Society of America (Allee, et al.,2000); Greene, et al. (1999); and NOAA Fisheries’ Our Living Oceans Habitat (Brown, 2000)project. In addition, habitat suitability modelling (Brown, et al., 2000) provides an approach forhabitat mapping based on environmental characteristics and the habitat requirements of aspecies or group of species.

The purpose of the Cowardin, et al., (1979) system, developed by the U.S. Department of theInterior, is to classify wetland and deepwater habitats for fish and wildlife. It is hierarchical, withfive systems at the top level: marine, estuarine, riverine, lacustrine, and palustrine (Figure 1).Each system contains 0-4 subsystems; each subsystem contains 1-8 classes, based onsubstrate, flooding regime, or vegetative form. Each class contains 2-7 subclasses, based ondominant substrate, or plant or animals forms. For example: Marine/subtidal/rockbottom/bedrock. Modifiers (e.g., based on chemistry or anthropogenic impacts) can be applied toclasses and subclasses.

The purpose of the Greene, et al., (1999) system is to characterize deepwater habitats ofvertebrates and invertebrates for establishment of marine reserves and for identifying EFH. It isan elaboration of a component of the Cowardin, et al., (1979) system, but is applied only to deepseafloor habitats. Four scales of habitat are recognized: megahabitat (>1 km), mesohabitat(10 m-1 km), macrohabitat (1-10 m), and microhabitat (cm). The hierarchy consists of system

Page 79: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

78

(marine benthic only), subsystem (mega- and mesohabitats based on physiography and depth),class (meso- and macrohabitats based on seafloor morphology), and subclass (macro- andmicrohabitats based on substratum textures or slope). Modifiers are included, based on bottommorphology, deposition, and texture; and/or on physical, chemical, biological, or anthropogenicprocesses.

The purpose of the system under development by NOAA Fisheries and the Ecological Society ofAmerica (Allee, et al., 2000) is to provide a framework for a national habitat classification forinventorying and tracking habitat changes. Although EFH is not referenced in Allee, et al.,(2000), establishing a national system for classifying habitat would have obvious relevance forany national or regional program to protect and conserve EFH. The proposed systememphasizes functional links between ecosystem structure and biological communities. Itestablishes a 13-level hierarchy (Table 2). This system has never been fully defined or tested,but it may form the basis for a more completely defined system currently being developed by acontractor.

NOAA Fisheries’ Our Living Oceans program publishes national syntheses on key aspects offishery management. The information is technical, but the reports are targeted at a lay audienceof senior resource managers, politicians and their staffs, and the general public. The purpose isto communicate with these constituencies about the status of the resources for which NOAAFisheries bears responsibility, and to provide them with an assessment of management andscientific needs. Several editions of the Our Living Oceans report on living marine resourcestocks were published in the 1990s. The most recent (NOAA Fisheries, 1999) describes thestatus of 25 species units, which cover over 600 fish, marine mammal, and sea turtle stocks. Thereport also summarizes the degree of utilization for the relevant fisheries. One edition of an OurLiving Oceans report on economics has been published (NOAA Fisheries, 1996). A new OurLiving Oceans report on habitat is currently in development (Brown, 2000). The long-term visionis to publish these three reports on a rotating annual basis.

The Our Living Oceans report on habitat is still being designed. Unlike the reports on stocks andeconomics, there is no well-established conceptual framework or program of research andmonitoring generating data for such a report. Nonetheless, the concepts for developing andanalyzing the data for the Our Living Oceans Habitat report have been defined (Brown, 2000).The report will be based on two fundamental components of information:

1. Species use of habitat – a checklist of the major habitat types used for each species by lifestage;

2. Usable habitat – quantity and quality of available habitat;3. Compared to the historical maximum; and4. Trends over the most recent ten years.

The data to be gathered will be qualitative in nature based on the knowledge of experts in eachregion. There will be no spatial data or maps in the initial version of the national database,although case studies of small regions will be included to illustrate the use of spatial data in theOur Living Oceans context. The long-term vision is to evolve into a framework based on spatialdata, but the Our Living Oceans team recognized that there is not a sufficient amount of spatialdata available to develop a national report on marine habitat at this time.

One of the initial requirements for developing the Our Living Oceans Habitat report is to develop ahabitat classification system. The current draft of this system is a six-level hierarchy. The toplevel consists of five broad categories: Fresh Water, Estuary, Nearshore, Offshore, and OffshoreIslands and Banks. Each habitat type within the hierarchy has its own unique numerical code(Table 3), which indicates where it fits within the hierarchy. This system is still being refined andreviewed by the Our Living Oceans team, including development of definitions for each habitattype.

Page 80: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

79

The habitat classification system for Our Living Oceans is intended for broad-scale tracking of thequantity and quality of habitat for the species for which NOAA Fisheries bears managementresponsibility. It is not highly detailed (e.g., there is only one habitat type for estuarine intertidalseagrass beds, with no finer-scale information on species composition, density, etc.). Therefore, ithas certain limitations. Because its intended use is to summarize habitat quantity for largeregions (e.g., Gulf of Maine), it has not been designed to handle mixed habitat types or small-scale patchiness.

Habitat Suitability Modelling

Habitat suitability modelling is a different approach to habitat classification and mapping. It has afairly long history, primarily in terrestrial and freshwater systems, which will not be reviewed here(e.g., U.S. Fish and Wildlife Service, 1980). Brown, et al., (2000) developed habitat suitabilityindex models for mapping habitat distributions for eight fish and invertebrate species in the Gulf ofMaine. These models classify habitat according to the habitat affinities of a species and lifestage. The mapping process, which requires use of a geographic information system, isillustrated in Figure 2. Each environmental variable in a model is plotted in a separate raster (i.e.,gridded) map. Each map is then reclassified to a model-specific suitability scale, which rangesfrom 1.0 for the most suitable range of the environmental variable, to 0.0 for ranges of thevariable in which the particular species or life stage does not occur. Then model calculations arethen made on a grid cell by grid cell basis. In Brown, et al., (2000), the models were thegeometric means of the suitability values of the variables in the model. More sophisticatedmodels could be used where data availability is sufficient to support their development.

The output of the above process is a map of habitat distribution. Figure 3 is an example forwinter flounder adults in Casco Bay during the summer. These maps should be interpreted at alevel appropriate for the information used in their development. In this case, spatial resolution islimited to 100 x 100 m grid cells. Temporal resolution is limited to season, because thetemperature and salinity maps used to calculate model outputs depict seasonal means. Prior toplotting the results, the model outputs were “binned” into categories of high, medium, and lowsuitability, and unsuitable. Comparisons for these categories were made among seasons,species and life stages. The models were not intended to be interpreted at small scales, such asby individual grid cells or at specific locations, such as a rock outcrop.

Conclusions

Habitat classification systems are designed for a particular purpose, and may depend on thescale and the organisms of interest. Developing a hierarchical system allows the flexibility tomove up or down in specificity and scale. Links among species, ecosystem function, and habitatsshould be clearly recognized. Many systems have been proposed and used for particularapplications. Developing a system specifically for maintaining a diversity of benthic habitat typeson the Scotian Shelf will require development of a classification system appropriate to the task. Itmay be possible to adapt an existing system, which would have the advantage of consistencywith other standards.

References

Allee, R.J., and 11 co-authors. 2000. Marine and Estuarine Ecosystem and Habitat Classification.NOAA Tech. Memo. NMFS-F/SPO-43. NOAA/National Marine Fisheries Service, Officeof Habitat Conservation, Silver Springs, Maryland, USA; 43p.

Brown, S.K. 2000. Draft Revised Outline: Our Living Oceans Habitat. December 26, 2000.NOAA/National Marine Fisheries Service, Silver Springs, Maryland, USA; 28p.

Brown, S.K., K.R. Buja, S.H. Jury, M.E. Monaco, and A. Banner. 2000. Habitat Suitability IndexModels for Eight Fish and Invertebrate Species in Casco and Sheepscot Bays, Maine. N.Amer. J. Fish. Manag. 20:408-435.

Page 81: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

80

Cowardin, L.M., V. Carter, F.C. Golet, and E.T. LaRoe. 1979. Classification of Wetlands andDeepwater Habitats of the United States. U.S. Department of the Interior ReportFWS/OBS-79/31. Available at http:://www.nwi.fws.gov/classifman/classman9.html.

Greene, H.G., and 7 co-authors. 1999. A Classification Scheme for Deep Seafloor Habitats.Oceanologica Acta 22(6):663-678.

NOAA Fisheries. 1996. Our Living Oceans. The economic status of U.S. fisheries, 1996. U.S.Department of Commerce; NOAA Tech. Memo. NMFS-F/SPO-22; 130p.

NOAA Fisheries. 1999. Our Living Oceans. Report on the status of U.S. living marine resources,1999. U.S. Department of Commerce; NOAA Tech. Memo. NMFS-F/SPO-41; 301p.

U.S. Fish and Wildlife Service. 1980. Habitat as a Basis for Environmental Assessment. U.S. Fishand Wildlife Service, Report 101 ESM. Fort Collins, Colorado, USA.

Page 82: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

81

Table 1. Major mandates that establish NOAA’s responsibilities for marine habitat.

Magnuson-Stevens Fishery Conservation and Management Act (also termed the SustainableFisheries Act) – establishes the regulatory framework for federally managed fisheries, includingEssential Fish Habitat requirements.

Endangered Species Act – requires identification and conservation of Critical Habitat for speciesat high risk of extinction. NOAA is responsible for fish, invertebrates, marine mammals, seaturtles, and one species of sea grass.

National Marine Sanctuaries Act – establishes marine sanctuaries in U.S. waters. Managementplans include conserving and protecting habitat.

Executive Order #13089 - establishes federal policy on coral reef protection, including mappingand conservation. NOAA and the U.S. Department of the Interior are the federal agenciesleading the Coral Reef Task Force.

Executive Order #13158 - establishes federal policy on marine protected areas (MPAs) as ameans for managing and conserving living marine resources and their habitats. NOAA leads theMPA Initiative, in cooperation with the U.S. Department of the Interior and many other partners.

Table 2. Thirteen-level NOAA Fisheries/Ecological Society of America habitat classificationhierarchy (Allee, et al., 2000).

1. Life Zone - climate2. Water/Land3. Marine/Estuarine/Freshwater4. Continental/Oceanic5. Bottom/Water Column6. Depth - shelf/slope/abyssal7. Regional Wave/Wind Energy8. Hydrogeomorphic/Earthform Features9. Hydrodynamic Features10. Photic/Aphotic11. Geomorphic Types/Topography12. Substratum/Eco-Type13. Local Modifiers and Eco-Units

Table 3. Example for Our Living Oceans habitat classification system, showing a segment forbottom habitats of the nearshore intertidal zone. The numbers are the numerical codefor each habitat type.

3 Nearshore (31 Supratidal Zone…) 32 Intertidal Zone 321 Bottom 3213 Vegetated bottom 32131 Rooted vascular 321311 Seagrass beds 32132 Algal beds 321321 Macroalgae (e.g., Fucus) 321322 Microalgae (e.g., calcareous green algae) 32133 Marine moss 32134 Emergent wetland

Page 83: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

82

System Subsystem Class

Marine Subtidal

Intertidal

Tidal

Lower Perennial

Upper Perennial

Intermittent

Limnetic

Littoral

Riverine

Lacustrine

Palustrine

Estuarine

WET

LAN

DS

AN

D D

EEPW

ATER

HA

BIT

ATS

Subtidal

Intertidal

Rock bottomUnconsolidated bottomAquatic bedReef

Aquatic bedReefRocky shoreUnconsolidated shore

Rock bottomUnconsolidated bottomAquatic bedReef

Aquatic bedReefStream bedRocky shoreUnconsolidated shoreEmergent wetlandScrub-Shrub wetlandForested wetland

Rock bottomUnconsolidated bottomAquatic bedStreambedRocky shoreUnconsolidated shoreEmergent wetland

Rock bottomUnconsolidated bottomAquatic bedRocky shoreUnconsolidated shoreEmergent wetland

Rock bottomUnconsolidated bottomAquatic bedRocky shoreUnconsolidated shore

StreambedRock bottomUnconsolidated bottomAquatic bed

Rock bottomUnconsolidated bottomAquatic bedRocky shoreUnconsolidated shoreEmergent wetland

Rock bottomUnconsolidated bottomAquatic bedUnconsolidated shoreMoss-lichen wetlandEmergent wetlandScrub-shrub wetlandForested wetland

Figure 1. Cowardin, et al., (1979) classification hierarchy.

Page 84: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

83

Figure 2. The process for classifying and mapping habitat using the habitat suitability indexmodels of Brown, et al., (2000).

Page 85: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

84

Figure 3. Habitat suitability index map for winter flounder adults in Casco Bay, Maine, during thesummer (Brown, et al., 2000).

Page 86: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

85

WORKSHOP OUTPUTS

Page 87: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

86

OUTPUTS OF WORKSHOP

Following the presentation of case studies and research results from several participants of theworkshop, a plenary discussion was held. The results presented below are mainly the commentsof workshop participants. They provide advice and direction to DFO as they proceed to developthe classification. The views outlined here do not necessarily represent official policy ormanagement position of DFO.

Day 1: Discussion Among Participants on the Goals of the Workshop

Workshop participants agreed that the primary goal is to define a marine habitat classificationframework that spans a range of organizational levels and spatial scales and that is suitable fordevelopment of integrated management plans for large ocean management units, and forplanning to preserve marine biodiversity.

The workshop outputs are not intended for the management of exploited fish stocks or fisheries.Although the outputs of the workshop are primarily of a scientific nature, the outcomes will be ofgreat interest to Scotian Shelf resource user groups (i.e. stakeholders) because of their directimplications for management by regulation and zoning. It was found necessary to ensure thatstakeholders are included as fully as they wish to be in the development of the classificationframework, and that the scientific basis for the resulting scheme is presented to stakeholders in amanner that is clear, defensible, and consistent with best practice elsewhere.

Consensus from the Participants on the Ocean Domain to be Considered

It was generally agreed that the large marine ecosystem of the Scotian Shelf to be classifiedextends from:

1. the middle of Laurentian Channel to the North, corresponding to regional boundary line, tothe border with the U.S. on the Georges Bank and in the Gulf of Maine to the South (explicitlyincluding the Bay of Fundy); and

2. the 50m depth contour closest to the coasts of Nova Scotia and New Brunswick (inshore) tothe base of the continental slope (1200m depth contour offshore).

These boundaries, although arbitrary in detail, reflect real and significant ecological separationsor transition zones, except in the case of the U.S. border (which is entirely political and cuts areasof demonstrable ecological integrity).

The classification framework to be developed should allow downstream compatibility with futureschemes to be developed for coastal areas and the U.S. areas of the Gulf of Maine and GeorgesBank (particular emphasis should be placed on joint development with U.S. counterparts in thelater case). Compatibility with mapping schemes in overlapping, conjoint and adjacentecosystems was thus seen as essential. Compatibility of the Scotian Shelf classification schemewith other classifications developed at larger scales and for non-adjacent ecosystems was foundnot essential, but desirable.

The Pelagic and Benthic communities of the Scotian Shelf are intimately and inseparably linkedby physical and ecological processes, and that a scientifically defensible benthic classificationscheme can not be developed in isolation from parameters of the overlying water column. Theworkload and available resources however may not allow simultaneous benthic and pelagicclassification activities.

Thus, at present time, it is possible to classify the benthic ecosystems of the Scotian Shelf in thecontext of pelagic ecosystem attributes, and to follow with the pelagic classification exercise intimely fashion, being prepared to revise the benthic classification in light of that futureclassification.

Page 88: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

87

Following the overview of workshop goals and lessons learned from other classification schemes,a series of nine (9) framework recommendations prepared and previously circulated by theconsultant were presented in plenary and discussed until consensus was reached on the secondday of the workshop.

Framework Recommendation #1

Development of a management oriented model for benthic classification that will:

1. Support development of an enforceable zoning scheme.2. Will guide human activities with the potential to reduce or degrade marine biodiversity.3. Support implementation of integrated oceans management.4. Improve understanding of marine ecosystems.5. Not place unrealistic requirements for data, time and funding.

It is necessary to apply a scientific approach to the process of habitat classification and mapping,recognizing the need to educate managers and resource users about the scientific basis andprocedures used for classification. The managers should set forth clear and realistic goals forscientific research on the topic of ecosystem classification. Both managers and scientists have tobe pro-active in this endeavour.

It is necessary to devise and assign metrics of accuracy and precision (i.e. uncertainty) to theindicators of benthic ecosystem type and their spatial boundaries. This was judged to be essentialif the classification scheme is to produce maps of benthic ecosystems that are useful formanagement decision support.

The development of a benthic classification scheme for the Scotian Shelf is not a “one-shot”activity. It must be an adaptive, continuously evolving process of hypothesis generation,experimentation, testing and revision. This workshop is seen as a starting point, not an end pointin the process.

Framework Recommendation #2

Set classification scales by capacity for compliance and enforcement such that:

1. The domain is defined by Ocean Management Area (OMA).2. Grain defined by the Minimum Manageable Unit (MMU).3. The Minimum Manageable Unit should exceed size of Minimum Mapable Unit.4. The Minimum Manageable Unit does not necessarily have to correspond to Minimum

Ecological Unit (MEU).

It is recognized that the actual size of the minimum management unit for the Scotian Shelf is yetto be determined, and it may not be the same size for all management strategies.

Framework Recommendation #3

Classify HIERARCHICALLY and zone ADAPTIVELY from the upper levels recognizing:

1. That hierarchies of biological organization and ecosystem function map less cleanly ontospatial-temporal scales than do physical-chemical processes.

2. Hierarchies of physical-chemical process to classify and map benthic habitat.3. Hierarchies of ecosystem structure and function sparingly within narrow ranges where

boundaries between levels are clear.

Spatial resolution in classification scheme is defined by the availability of data, whereas the sizesof management units depend on the nature of managed activity.

Page 89: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

88

There is an hierarchy of management units, some of which are rather small (e.g. point sourceimpacts such as drilling rigs, known locations of rare and valuable, sessile organisms). Bufferzones around such small areas, depending on the kind of activity would bring the size of themanagement unit up to the tens or km2. Moving down the hierarchies of spatial scale for bothbenthic mapping and ocean management increases the uncertainty and costs exponentially, andit is recommended that such decisions would be made on a “need-to-know” basis.

A spatial hierarchy of management units matches a spatial hierarchy of mapable units, bothhaving different levels of uncertainty, depending on the technologies used. Remote sensingtechnologies can map oceanographic features down to 10km2 with a precision of 1 km2, andseabed units down to 100m2 with a precision of 1m2. The main challenge here is downscalingfrom the synoptic view to the MEU.

Benthic sampling technologies can map macrofaunal communities down to 10m2, but with a lowprecision on the order of 10,000m2 because of the errors of interpolation among samples. Themain challenge here is up-scaling from the sample unit to the habitat and shelf scales.

Framework Recommendation #4

Use the benthic assemblages as the Minimum Ecological Unit (MEU):

1. Communities or species assemblages should be considered as indicators of environmentalconditions.

2. Benthic habitat associated with assemblages of species should be considered as theminimum mapable unit (MMU).

3. Close correspondence of benthic communities to physically-defined habitats should besupported by empirical data.

Benthic habitat explicitly includes structuring biotic components (e.g. corals), as well as physicalstructure and biophysical attributes of the overlying water column. Biological community respondsto ambient physical factors and simultaneously modifies them. Additionally the members ofbiological community have unique and often strong influence of each other.

Causal and statistical relationships established between the distribution and abundance ofbenthic organisms (such as individuals, populations or species) and the structure of the benthichabitat are highly variable, and often of low predictive power. However, the relationships betweenstructural and particularly functional attributes of enduring benthic communities and the physicalattributes of their characteristic habitats are more robust, and thus potentially more useful for thepurposes of this exercise. The use of spatially repeated functional groupings of organisms wasrecommended for identifying mapable benthic habitats.

Framework Recommendation #5

Incorporate Oceanographic and Trophodynamic processes explicitly where:

1. Water column structure and processes or pelagic communities can not be consideredseparately from benthic communities, as they exert dominant controls, such as;• regulate environment,• create disturbance regime,• deliver food and recruits, and• cycle energy and materials.

2. Including nearshore areas especially in the Bay of Fundy, but not as strongly off Southernand Eastern Nova Scotia because of oceanographic separation of coastal and shelf watermasses.

Page 90: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

89

Pelagic and Benthic communities of the Scotian Shelf are intimately an inseparably linked byphysical and ecological processes, and that a scientifically defensible benthic classificationscheme can not be developed in isolation from parameters of the overlying water column.

Framework Recommendation #6

Incorporate the role of HISTORY in classification of benthic habitats:

1. Distribution and abundance of organisms not fixed in space-time and may vary at scalessmaller-faster than MMUs.

2. Variation in oceanographic conditions (e.g. organic supply to benthos) is cumulative and theeffects are often lagged.

3. Disturbance frequency and intensity exert strong controls on benthic communities, it mayhave physical or biotic origin.

4. Need to distinguish and quantify natural (e.g. sediment mobility, species interactions) andanthropogenic disturbance rates.

Mapping natural and anthropogenic disturbance is essential to the implementation ofmanagement strategies, and therefore such information must be explicitly identified andincorporated with the benthic classification scheme.

Framework Recommendation #7

Operationalize classification by MAPPING enduring habitat units:

1. Habitat units should be defined by;• marine environment,• substratum,• food supply, and• disturbance regime.

2. Define benthic community types based on theoretical and empirical relationships with thesehabitat attributes.

3. Scale-up from verified and calibrated areas to entire shelf using synoptic maps throughIteration, improvement and cross-validation.

4. Sampled biota and measured physical variates may be classified a posteriori.5. Multivariate tools should be used for classification because they are powerful classifiers even

in the absence of mechanistic understanding.6. Pattern description is suitable for habitat mapping.

The characteristic scale of minimum mapped units (i.e. the grain) of the benthic habitatclassification must be larger than the characteristic size of habitat patches that are subject to highfrequency (i.e. faster than decadal period) variation in their location or boundaries.

Development, implementation and adaptive revision of zoning-based management strategiesderived from the benthic habitat maps could range from years to a decade and therefore shouldbe attempted as soon as possible.

Framework Recommendation #8

Map benthic habitats first on the basis of SHELF-SCALE SYNOPSES of:

1. Oceanography: water masses, circulations, regimes (4)2. Physiography: topographic hierarchy (2 levels, 8 classes)3. Substratum (2 levels, 7 classes)4. Commercial fish and by-catch5. Invertebrates6. Rare, long-lived and structuring species

Page 91: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

90

7. Finer-grained, continuos variables should be added if available (e.g. U*)8. Second on the basis of sampled distributions of biota

Detailed description of initial benthic habitat mapping framework for the Scotian Shelf ispresented at the end of this document.

Framework Recommendation #9

Relate HABITAT COMPLEXITY to biodiversity:

1. Biodiversity is well-related to complexity of environment;• Architectural complexity of substrates increases alpha diversity, and• Variability of habitat types within an area increases beta diversity.

2. Use habitat attributes and landscape structure as proxys, surrogates or indicators for diversityof benthic organisms.

3. Establish reference biodiversity areas (such as potential MPAs).

Management goal of preserving benthic biodiversity on the Scotian Shelf is not restricted tosimply identifying and protecting those areas identified as having the maximum biodiversity. Itentails protecting distinct habitats that may support both high and low diversities of structure andlife forms.

Day2: Proposed Benthic Habitat Classification and Mapping Framework

In the final activity of the workshop, the group discussed in plenary a proposed, minimum initialset of shelf habitat zones, features, variables and metrics that would be compiled and mapped toproduce a draft framework. This will then be circulated to the scientific community andstakeholder groups for criticism and revision as part of the second phase of the project.

The following is a description of a proposed fully hierarchical classification scheme where alllevels are nested in higher levels and no attribute at one level exists at another level.

Level 1: Oceanographic Domains

• Six (6) domains were recognized:- North-eastern Scotian Shelf,- South-eastern Scotian Shelf,- Central Scotian Shelf,- Western Scotian Shelf,- Scotian Shelf Edge, and- Gulf of Maine (including Georges Bank and Bay of Fundy).

• Five (5) classifying attributes were recommended to map these domains:

1. Critical depths (related to the essential ecological processes of food supply, growth,disturbance, reproduction/dispersal/recruitment, and mortality).- Metrics / Variates (scaled): Pleistocene still-stand depth/z; mixed layer depth/z; photic

layer depth/z to a precision of 1m (derived from digital bathymetry).2. Thermal regime as epibenthic (bottom) seawater temperature (causally related to the

survival and growth of benthic biota).- Metrics / Variates: Min. and Max. annual temperature to a precision of 1oC (from

numerical models).3. Current regime as epibenthic (near-bed) water current velocity gradient (causally related to

the disturbance, dispersal and food supply).- Metrics / Variates: typical value of U* to a precision of 0.01 ms-2 (from numerical models).

4. Productivity regime as persistent plankton features (causally related to food supply).

Page 92: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

91

- Metrics / Variates: Season and depth-integrated annual primary productivity of the watermass to a precision of 50 gC.m-2.y-1 (from remote sensing and analytical models; mayhave to be satisfied with Chlorophyll-a concentration).

5. Disturbance regime as storm intensity and frequency (related to food supply, dispersal andmortality).- Metrics / Variates: Annual probability of sand resuspension and transport for 24h or

longer to a precision of 5 d.y-1 (from numerical models).

Levels 2-3: Seabed Domains Within Oceanographic Domains

• Four (4) Physiographic domains were recognized:- Inner Scotian Shelf,- Middle Scotian Shelf- Outer Scotian Shelf- Scotian Slope

• Within which eight (8) Morphological domains were recognized:- Banks- Bank Flanks- Basins- Saddles- Intermediate continuum (connecting those above, corresponding to “Valleys and Plains”)- Canyons- Upper Slope (Slope-Shelf transition)- Lower slope (Slope –Ocean basin transition)

• Within which four (4) scales of Seabed texture were recognized:

1. Topographic roughness as measured by the second derivative of depth at a spatial scale of1-100 km (related to dispersal, recruitment and diversity).- Metrics / Variates: z* to a precision of 1 m-2.

2. Surficial complexity as measured by the rugosity index or surface fractal dimention at aspatial scale of 100 – 1,000 m (related to survival, recruitment and diversity).- Metrics / Variates: ratio of actual to vertically projected length, or F, to a precision of 0.1,

fractal dimention to a precision of 0.01.3. Bio-structural complexity as measured by the aspect ratio of biogenic structural elements

at a spatial scale of 1-100 m (related to (growth, survival and recruitment).- Metrics / Variates: ratio of mean element height to inter-element distance to a precision of

0.1.4. Particle roughness as measured by the grain characteristics at a spatial scale of 10 – 1,000

cm (related to feeding, growth, survival, recruitment and diversity).5. Five types of substratum are recognized;

1. Rock (outcrop or bedrock, which is sub-divided into hard and soft rock),2. Gravel,3. Sand,4. Sand/Mud, and5. Mud.

- Metrics/Variates:1. Mean grain size to a precision of 0.001 mm or Phi to a precision of 0.1 (related to

rugosity and pore size).2. Skewness to a precision of 0.01 (related to sediment erosion and resuspension).3. Cohesiveness to a precision of 0.1 dynes (related to biocementation).

Page 93: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

92

Level 4: Biological Communities

Benthic macro- and megafaunal assemblages as described from optical underwater observationsand physical grab and dredge samples allow for synoptic description of animal-habitatassociations on the seabed.

Metrics/Variates:1. Within-habitat species diversity,2. Frequency of occurrence, biomass, abundance of individual species,3. Degree of association between species and physical habitat,4. Dominant megafaunal species, and5. List of associated species for each habitat type.

Maps of the distribution and abundance of benthic assemblages at shelf-scales, as well as mapsof other organisms or characteristic assemblages of organisms at smaller scales (e.g. withinmorphological domains or areas of uniform seabed texture), will play key roles in thedevelopment of the framework classification of benthic ecosystem types on the Scotian Shelf.Some of these data sets exist already, while others will be identified as priority topics for futureresearch and experimentation. Several groups of foundation organisms were recognized as well-enough known by their fundamental niches, ecological roles and distribution and abundance onthe Scotian shelf may be mapped at this stage. These may be surf clams, scallops and sanddollars, demersal fish, gorgonian and scleractinian corals.

The uses of these distribution maps of biological components of the Scotian Shelf ecosystem are:

1. To demonstrate and quantify empirical relationships between physical attributes of thehabitats and their biological diversity using rigorous parametric and multivariate statisticaltools.

2. To test and verify predictions of the distribution and abundance of organisms based onextrapolations from the benthic habitat maps.

3. To revise and improve the accuracy and detail of the proposed classification system to thepoint where the proxies, indicators and reference points provide the best possible support formanagement decisions regarding the preservation of biodiversity on the Scotian Shelf.

The proposed approaches stated here will be taken in consideration by DFO in developingbenthic habitat classification.

Page 94: Proceedings of a Benthic Habitat Classification Workshop ... · in this report may be factually incorrect or mis- ... System for Multiple Uses ... Revoir les approches existantes

Maritimes Region Benthic Habitat Classification Workshop

93

BENTHIC WORKSHOP ATTENDEES

DFO Maritimes

Paul Boudreau - MESD/Habitat EcologistDon Gordon - MESD/Research AssistantEllen Kenchington - Science/Research ScientistBob O’Boyle - Science/RAP CoordinatorBob Rutherford - OCMD/Section HeadDerek Fenton - OCMD/Policy AdvisorJoe Arbour - OCMD/Division ManagerPaul Macnab - OCMD/Oceans Act CoordinationsGlen Herbert - OCMD/Oceans Act CoordinationsScott Coffen-Smout - OCMD/Oceans Act CoordinationsHeather Breeze - OCMDJennifer Hackett - OCMD/Oceans Management CoordinatorDenise McCullough - OCMD/Community Initiatives OfficerBarry Jones - OCMD/Section HeadGlen Harrison - Science /Research ScientistMaxine Westhead - OCMD/Oceans Biologist/Bay of FundyJason Naug - OCMD/Oceans Policy Officer

Non-DFO/Federal Government

Brian Todd - Natural Resources Canada/GSC AtlanticVladimir Kostylev - GSC/DFODick Pickrill - Natural Resources Canada/GSC AtlanticGordon Fader - Natural Resources Canada/GSC AtlanticKyle Penney - Department of National Defence/Formation Environment

Non-DFO/Government

Derek Davis - MIDI Marine Invertebrates Diversity Initiative/ChairBruce Hatcher - Canadian Fishery Consultants Ltd.Brian Giroux - SF Mobile Gear Fishermen’s AssociationLea-Ann Henry - Dalhousie University/Department of BiologyMartin Willison - Dalhousie UniversityPat Stewart - Envirosphere Consultants Ltd.Brad Barr - NOAA-National Oceanic and Atmospheric AdministrationSteve Brown - NOAA-National Oceanic and Atmospheric AdministrationMarty King - Dalhousie University/StudentJayne Roma - MIDI/CoordinatorBruce Chapman - Fisheries Resource Conservation CouncilMark Butler - Ecology Action CentreNancy Witherspoon - Canadian Aquatic Environmental Consultants Inc.Susan Gass - Ecology Action CentreKim Doane - Nova Scotia Petroleum DirectorateAndre d’Entremont - Kerr-McGee Inc.